[
{
"path": ".gitignore",
"content": "attent*\nlog*\n*.pth\ncheckpoint\n1st\nallmetaclass\nbase*\noutput*\nsplit*\nresult*\n*.sh\n"
},
{
"path": "README.md",
"content": "## Meta R-CNN : Towards General Solver for Instance-level Low-shot Learning.\r\n\r\nCode for reproducing the results in the following paper, and the code is built on top of [jwyang/faster-rcnn.pytorch](https://github.com/jwyang/faster-rcnn.pytorch)\r\n\r\n**Meta R-CNN : Towards General Solver for Instance-level Low-shot Learning**\r\n\r\nXiaopeng Yan*,\r\nZiliang Chen*,\r\nAnni Xu, Xiaoxi Wang, \r\nXiaodan Liang,\r\nLiang Lin\r\n\r\nSun Yat-Sen University, Presented at *IEEE International Conference on Computer Vision [(ICCV2019)](http://iccv2019.thecvf.com/)*\t\r\n\r\n![](\"demo/metarcnn.png\"/)
\r\n\r\n### License\r\n\r\nFor Academic Research Use Only!\r\n\r\n### Requirements\r\n\r\n+ python packages\r\n\r\n + PyTorch = 0.3.1\r\n \r\n *This project can not support pytorch 0.4, higher version will not recur results.*\r\n\r\n + Torchvision >= 0.2.0\r\n\r\n + cython\r\n\r\n + pyyaml\r\n\r\n + easydict\r\n\r\n + opencv-python\r\n\r\n + matplotlib\r\n\r\n + numpy\r\n\r\n + scipy\r\n\r\n + tensorboardX\r\n\r\n You can install above package using ```pip```:\r\n\r\n ```sh\r\n pip install Cython easydict matplotlib opencv-python pyyaml scipy\r\n ```\r\n\r\n+ CUDA 8.0\r\n\r\n+ gcc >= 4.9\r\n\r\n### Misc\r\n\r\nTested on Ubuntu 14.04 with a Titan X GPU (12G) and Intel(R) Xeon(R) CPU E5-2623 v3 @ 3.00GHz.\r\n\r\n### Getting Started\r\n\r\nClone the repo:\r\n\r\n```\r\nhttps://github.com/yanxp/MetaR-CNN.git\r\n```\r\n\r\n### Compilation\r\n\r\nCompile the CUDA dependencies:\r\n\r\n```sh\r\ncd {repo_root}/lib\r\nsh make.sh\r\n```\r\nIt will compile all the modules you need, including NMS, ROI_Pooing, ROI_Crop and ROI_Align. \r\n\r\n### Data Preparation\r\n\r\nCreate a data folder under the repo,\r\n\r\n```sh\r\ncd {repo_root}\r\nmkdir data\r\n```\r\n**PASCAL_VOC 07+12**: Please follow the instructions in [py-faster-rcnn](https://github.com/rbgirshick/py-faster-rcnn#beyond-the-demo-installation-for-training-and-testing-models) to prepare VOC datasets. Actually, you can refer to any others. After downloading the data, create softlinks in the folder data/.\r\n\r\nplease download the three base classes [splits](https://pan.baidu.com/s/11IxGujTTegLEXFsaiohV_Q)[[GoogleDrive](https://drive.google.com/drive/folders/14gtxnxWokk3eO6Oe5SrEG6_R9Dt6efT8?usp=sharing)] and put them into VOC2007 and VOC2012 ImageSets/Main dirs.\r\n\r\n### Training\r\nWe used [ResNet101](https://www.dropbox.com/s/iev3tkbz5wyyuz9/resnet101_caffe.pth?dl=0) pretrained model on ImageNet in our experiments. Download it and put it into the data/pretrained_model/.\r\n\r\nfor example, if you want to train the first split of base and novel class with meta learning, just run:\r\n\r\n#### the first phase\r\n```sh\r\n$>CUDA_VISIBLE_DEVICES=0 python train_metarcnn.py --dataset pascal_voc_0712 --epochs 21 --bs 4 --nw 8 --log_dir checkpoint --save_dir models/meta/first --meta_type 1 --meta_train True --meta_loss True \r\n```\r\n#### the second phase\r\n```sh\r\n$>CUDA_VISIBLE_DEVICES=0 python train_metarcnn.py --dataset pascal_voc_0712 --epochs 30 --bs 4 --nw 8 --log_dir checkpoint --save_dir models/meta/first --r True --checksession 1 --checkepoch 20 --checkpoint 3081 --phase 2 --shots 10 --meta_train True --meta_loss True --meta_type 1\r\n```\r\n### Testing\r\n\r\nif you want to evaluate the performance of meta trained model, simply run:\r\n```sh\r\n$>CUDA_VISIBLE_DEVICES=0 python test_metarcnn.py --dataset pascal_voc_0712 --net metarcnn --load_dir models/meta/first --checksession 10 --checkepoch 30 --checkpoint 111 --shots 10 --meta_type 1 --meta_test True --meta_loss True --phase 2\r\n```\r\n\r\nwe provide the part models with meta training and without meta training in the following:\r\n[Meta Models](https://pan.baidu.com/s/1N3PW9WTi82lbdURNAz7EFA)[[GoogleDrive](https://drive.google.com/file/d/19gapxklxKCwYIyGszOMhQKNDqYOLeubn/view?usp=sharing)] and [WoMeta Models](https://pan.baidu.com/s/1GkjUJmaOaEWzh3z2fs7ieA)[[GoogleDrive](https://drive.google.com/file/d/1G6xYH9M_bAAqUec1ARufv0ELi_pd7ERj/view?usp=sharing)]\r\n\r\n### Citation\r\n\r\n```\r\n@inproceedings{yan2019meta,\r\n title={Meta r-cnn: Towards general solver for instance-level low-shot learning},\r\n author={Yan, Xiaopeng and Chen, Ziliang and Xu, Anni and Wang, Xiaoxi and Liang, Xiaodan and Lin, Liang},\r\n booktitle={Proceedings of the IEEE International Conference on Computer Vision},\r\n pages={9577--9586},\r\n year={2019}\r\n}\r\n```\r\n\r\n### Contact\r\n\r\nIf you have any questions about this repo, please feel free to contact [yanxp3@mail3.sysu.edu.cn](mailto:yanxp3@mail3.sysu.edu.cn).\r\n"
},
{
"path": "_init_paths.py",
"content": "import os.path as osp\nimport sys\nimport os\n\nif os.listdir('data/cache/'):\n os.system('rm data/cache/*')\n\ndef add_path(path):\n if path not in sys.path:\n sys.path.insert(0, path)\n\nthis_dir = osp.dirname(__file__)\n\n# Add lib to PYTHONPATH\nlib_path = osp.join(this_dir, 'lib')\nadd_path(lib_path)\n\ncoco_path = osp.join(this_dir, 'data', 'coco', 'PythonAPI')\nadd_path(coco_path)\n\nvg_path = osp.join(this_dir, 'data', 'vgapi')\nadd_path(vg_path)\n"
},
{
"path": "cfgs/res101_ms.yml",
"content": "EXP_DIR: res101\nTRAIN:\n HAS_RPN: True\n BBOX_NORMALIZE_TARGETS_PRECOMPUTED: True\n RPN_POSITIVE_OVERLAP: 0.7\n RPN_BATCHSIZE: 256\n PROPOSAL_METHOD: gt\n BG_THRESH_LO: 0.0\n DISPLAY: 20\n BATCH_SIZE: 128\n WEIGHT_DECAY: 0.0001\n MAX_SIZE: 1000\n SCALES: [600]\n DOUBLE_BIAS: False\n RCNN_BBOX_WEIGHT: 1\nTEST:\n SCALES: [600]\n HAS_RPN: True\nPOOLING_SIZE: 7\nPOOLING_MODE: align\nCROP_RESIZE_WITH_MAX_POOL: False\n"
},
{
"path": "cfgs/res50.yml",
"content": "EXP_DIR: res50\nTRAIN:\n HAS_RPN: True\n BBOX_NORMALIZE_TARGETS_PRECOMPUTED: True\n RPN_POSITIVE_OVERLAP: 0.7\n RPN_BATCHSIZE: 256\n PROPOSAL_METHOD: gt\n BG_THRESH_LO: 0.0\n DISPLAY: 20\n BATCH_SIZE: 128\n WEIGHT_DECAY: 0.0001\n MAX_SIZE: 1000\n SCALES: [600]\n DOUBLE_BIAS: False\n RCNN_BBOX_WEIGHT: 1\nTEST:\n SCALES: [600]\n HAS_RPN: True\nPOOLING_SIZE: 7\nPOOLING_MODE: align\nCROP_RESIZE_WITH_MAX_POOL: False\n"
},
{
"path": "lib/datasets/VOCdevkit-matlab-wrapper/get_voc_opts.m",
"content": "function VOCopts = get_voc_opts(path)\n\ntmp = pwd;\ncd(path);\ntry\n addpath('VOCcode');\n VOCinit;\ncatch\n rmpath('VOCcode');\n cd(tmp);\n error(sprintf('VOCcode directory not found under %s', path));\nend\nrmpath('VOCcode');\ncd(tmp);\n"
},
{
"path": "lib/datasets/VOCdevkit-matlab-wrapper/voc_eval.m",
"content": "function res = voc_eval(path, comp_id, test_set, output_dir)\n\nVOCopts = get_voc_opts(path);\nVOCopts.testset = test_set;\n\nfor i = 1:length(VOCopts.classes)\n cls = VOCopts.classes{i};\n res(i) = voc_eval_cls(cls, VOCopts, comp_id, output_dir);\nend\n\nfprintf('\\n~~~~~~~~~~~~~~~~~~~~\\n');\nfprintf('Results:\\n');\naps = [res(:).ap]';\nfprintf('%.1f\\n', aps * 100);\nfprintf('%.1f\\n', mean(aps) * 100);\nfprintf('~~~~~~~~~~~~~~~~~~~~\\n');\n\nfunction res = voc_eval_cls(cls, VOCopts, comp_id, output_dir)\n\ntest_set = VOCopts.testset;\nyear = VOCopts.dataset(4:end);\n\naddpath(fullfile(VOCopts.datadir, 'VOCcode'));\n\nres_fn = sprintf(VOCopts.detrespath, comp_id, cls);\n\nrecall = [];\nprec = [];\nap = 0;\nap_auc = 0;\n\ndo_eval = (str2num(year) <= ) | ~strcmp(test_set, 'test');\nif do_eval\n % Bug in VOCevaldet requires that tic has been called first\n tic;\n [recall, prec, ap] = VOCevaldet(VOCopts, comp_id, cls, true);\n ap_auc = xVOCap(recall, prec);\n\n % force plot limits\n ylim([0 1]);\n xlim([0 1]);\n\n print(gcf, '-djpeg', '-r0', ...\n [output_dir '/' cls '_pr.jpg']);\nend\nfprintf('!!! %s : %.4f %.4f\\n', cls, ap, ap_auc);\n\nres.recall = recall;\nres.prec = prec;\nres.ap = ap;\nres.ap_auc = ap_auc;\n\nsave([output_dir '/' cls '_pr.mat'], ...\n 'res', 'recall', 'prec', 'ap', 'ap_auc');\n\nrmpath(fullfile(VOCopts.datadir, 'VOCcode'));\n"
},
{
"path": "lib/datasets/VOCdevkit-matlab-wrapper/xVOCap.m",
"content": "function ap = xVOCap(rec,prec)\r\n% From the PASCAL VOC 2011 devkit\r\n\r\nmrec=[0 ; rec ; 1];\r\nmpre=[0 ; prec ; 0];\r\nfor i=numel(mpre)-1:-1:1\r\n mpre(i)=max(mpre(i),mpre(i+1));\r\nend\r\ni=find(mrec(2:end)~=mrec(1:end-1))+1;\r\nap=sum((mrec(i)-mrec(i-1)).*mpre(i));\r\n"
},
{
"path": "lib/datasets/__init__.py",
"content": "# --------------------------------------------------------\n# Fast R-CNN\n# Copyright (c) 2015 Microsoft\n# Licensed under The MIT License [see LICENSE for details]\n# Written by Ross Girshick\n# --------------------------------------------------------"
},
{
"path": "lib/datasets/coco.py",
"content": "# --------------------------------------------------------\n# Fast/er R-CNN\n# Licensed under The MIT License [see LICENSE for details]\n# Written by Ross Girshick and Xinlei Chen\n# --------------------------------------------------------\nfrom __future__ import absolute_import\nfrom __future__ import division\nfrom __future__ import print_function\n\nfrom datasets.imdb import imdb\nimport datasets.ds_utils as ds_utils\nfrom model.utils.config import cfg\nimport os.path as osp\nimport sys\nimport os\nimport numpy as np\nimport scipy.sparse\nimport scipy.io as sio\nimport pickle\nimport json\nimport uuid\n# COCO API\nfrom pycocotools.coco import COCO\nfrom pycocotools.cocoeval import COCOeval\nfrom pycocotools import mask as COCOmask\n\nclass coco(imdb):\n def __init__(self, image_set, year):\n imdb.__init__(self, 'coco_' + year + '_' + image_set)\n # COCO specific config options\n self.config = {'use_salt': True,\n 'cleanup': True}\n # name, paths\n self._year = year\n self._image_set = image_set\n self._data_path = osp.join(cfg.DATA_DIR, 'coco'+self._year)\n # load COCO API, classes, class <-> id mappings\n self._COCO = COCO(self._get_ann_file())\n cats = self._COCO.loadCats(self._COCO.getCatIds())\n self._classes = tuple(['__background__'] + [c['name'] for c in cats])\n self._class_to_ind = dict(list(zip(self.classes, list(range(self.num_classes)))))\n self._class_to_coco_cat_id = dict(list(zip([c['name'] for c in cats],\n self._COCO.getCatIds())))\n self._image_index = self._load_image_set_index()\n # Default to roidb handler\n self.set_proposal_method('gt')\n self.competition_mode(False)\n\n # Some image sets are \"views\" (i.e. subsets) into others.\n # For example, minival2014 is a random 5000 image subset of val2014.\n # This mapping tells us where the view's images and proposals come from.\n self._view_map = {\n 'minival2014': 'val2014', # 5k val2014 subset\n 'valminusminival2014': 'val2014', # val2014 \\setminus minival2014\n 'test-dev2015': 'test2015',\n 'valminuscapval2014': 'val2014',\n 'capval2014': 'val2014',\n 'captest2014': 'val2014'\n }\n coco_name = image_set + year # e.g., \"val2014\"\n self._data_name = (self._view_map[coco_name]\n if coco_name in self._view_map\n else coco_name)\n # Dataset splits that have ground-truth annotations (test splits\n # do not have gt annotations)\n self._gt_splits = ('train', 'val', 'minival')\n\n def _get_ann_file(self):\n prefix = 'instances' if self._image_set.find('test') == -1 \\\n else 'image_info'\n return osp.join(self._data_path, 'annotations',\n prefix + '_' + self._image_set + self._year + '.json')\n\n def _load_image_set_index(self):\n \"\"\"\n Load image ids.\n \"\"\"\n image_ids = self._COCO.getImgIds()\n return image_ids\n\n def _get_widths(self):\n anns = self._COCO.loadImgs(self._image_index)\n widths = [ann['width'] for ann in anns]\n return widths\n\n def image_path_at(self, i):\n \"\"\"\n Return the absolute path to image i in the image sequence.\n \"\"\"\n return self.image_path_from_index(self._image_index[i])\n\n def image_id_at(self, i):\n \"\"\"\n Return the absolute path to image i in the image sequence.\n \"\"\"\n return self._image_index[i]\n\n def image_path_from_index(self, index):\n \"\"\"\n Construct an image path from the image's \"index\" identifier.\n \"\"\"\n # Example image path for index=119993:\n # images/train2014/COCO_train2014_000000119993.jpg\n if self._year == '2017':\n file_name = str(index).zfill(12) + '.jpg'\n elif self._year == '2014':\n file_name = ('COCO_' + self._data_name + '_' +\n str(index).zfill(12) + '.jpg')\n image_path = osp.join(self._data_path, 'images',\n self._data_name, file_name)\n assert osp.exists(image_path), \\\n 'Path does not exist: {}'.format(image_path)\n return image_path\n\n def gt_roidb(self):\n \"\"\"\n Return the database of ground-truth regions of interest.\n This function loads/saves from/to a cache file to speed up future calls.\n \"\"\"\n cache_file = osp.join(self.cache_path, self.name + '_gt_roidb.pkl')\n if osp.exists(cache_file):\n with open(cache_file, 'rb') as fid:\n roidb = pickle.load(fid)\n print('{} gt roidb loaded from {}'.format(self.name, cache_file))\n return roidb\n\n gt_roidb = [self._load_coco_annotation(index)\n for index in self._image_index]\n\n with open(cache_file, 'wb') as fid:\n pickle.dump(gt_roidb, fid, pickle.HIGHEST_PROTOCOL)\n print('wrote gt roidb to {}'.format(cache_file))\n return gt_roidb\n\n def _load_coco_annotation(self, index):\n \"\"\"\n Loads COCO bounding-box instance annotations. Crowd instances are\n handled by marking their overlaps (with all categories) to -1. This\n overlap value means that crowd \"instances\" are excluded from training.\n \"\"\"\n im_ann = self._COCO.loadImgs(index)[0]\n width = im_ann['width']\n height = im_ann['height']\n\n annIds = self._COCO.getAnnIds(imgIds=index, iscrowd=None)\n objs = self._COCO.loadAnns(annIds)\n # Sanitize bboxes -- some are invalid\n valid_objs = []\n for obj in objs:\n x1 = np.max((0, obj['bbox'][0]))\n y1 = np.max((0, obj['bbox'][1]))\n x2 = np.min((width - 1, x1 + np.max((0, obj['bbox'][2] - 1))))\n y2 = np.min((height - 1, y1 + np.max((0, obj['bbox'][3] - 1))))\n if obj['area'] > 0 and x2 >= x1 and y2 >= y1:\n obj['clean_bbox'] = [x1, y1, x2, y2]\n valid_objs.append(obj)\n objs = valid_objs\n num_objs = len(objs)\n\n boxes = np.zeros((num_objs, 4), dtype=np.uint16)\n gt_classes = np.zeros((num_objs), dtype=np.int32)\n overlaps = np.zeros((num_objs, self.num_classes), dtype=np.float32)\n seg_areas = np.zeros((num_objs), dtype=np.float32)\n\n # Lookup table to map from COCO category ids to our internal class\n # indices\n coco_cat_id_to_class_ind = dict([(self._class_to_coco_cat_id[cls],\n self._class_to_ind[cls])\n for cls in self._classes[1:]])\n\n for ix, obj in enumerate(objs):\n cls = coco_cat_id_to_class_ind[obj['category_id']]\n boxes[ix, :] = obj['clean_bbox']\n gt_classes[ix] = cls\n seg_areas[ix] = obj['area']\n if obj['iscrowd']:\n # Set overlap to -1 for all classes for crowd objects\n # so they will be excluded during training\n overlaps[ix, :] = -1.0\n else:\n overlaps[ix, cls] = 1.0\n\n ds_utils.validate_boxes(boxes, width=width, height=height)\n overlaps = scipy.sparse.csr_matrix(overlaps)\n return {'width': width,\n 'height': height,\n 'boxes': boxes,\n 'gt_classes': gt_classes,\n 'gt_overlaps': overlaps,\n 'flipped': False,\n 'seg_areas': seg_areas}\n\n def _get_widths(self):\n return [r['width'] for r in self.roidb]\n\n def append_flipped_images(self):\n num_images = self.num_images\n widths = self._get_widths()\n for i in range(num_images):\n boxes = self.roidb[i]['boxes'].copy()\n oldx1 = boxes[:, 0].copy()\n oldx2 = boxes[:, 2].copy()\n boxes[:, 0] = widths[i] - oldx2 - 1\n boxes[:, 2] = widths[i] - oldx1 - 1\n assert (boxes[:, 2] >= boxes[:, 0]).all()\n entry = {'width': widths[i],\n 'height': self.roidb[i]['height'],\n 'boxes': boxes,\n 'gt_classes': self.roidb[i]['gt_classes'],\n 'gt_overlaps': self.roidb[i]['gt_overlaps'],\n 'flipped': True,\n 'seg_areas': self.roidb[i]['seg_areas']}\n\n self.roidb.append(entry)\n self._image_index = self._image_index * 2\n\n def _get_box_file(self, index):\n # first 14 chars / first 22 chars / all chars + .mat\n # COCO_val2014_0/COCO_val2014_000000447/COCO_val2014_000000447991.mat\n file_name = ('COCO_' + self._data_name +\n '_' + str(index).zfill(12) + '.mat')\n return osp.join(file_name[:14], file_name[:22], file_name)\n\n def _print_detection_eval_metrics(self, coco_eval):\n IoU_lo_thresh = 0.5\n IoU_hi_thresh = 0.95\n\n def _get_thr_ind(coco_eval, thr):\n ind = np.where((coco_eval.params.iouThrs > thr - 1e-5) &\n (coco_eval.params.iouThrs < thr + 1e-5))[0][0]\n iou_thr = coco_eval.params.iouThrs[ind]\n assert np.isclose(iou_thr, thr)\n return ind\n\n ind_lo = _get_thr_ind(coco_eval, IoU_lo_thresh)\n ind_hi = _get_thr_ind(coco_eval, IoU_hi_thresh)\n # precision has dims (iou, recall, cls, area range, max dets)\n # area range index 0: all area ranges\n # max dets index 2: 100 per image\n precision = \\\n coco_eval.eval['precision'][ind_lo:(ind_hi + 1), :, :, 0, 2]\n ap_default = np.mean(precision[precision > -1])\n print(('~~~~ Mean and per-category AP @ IoU=[{:.2f},{:.2f}] '\n '~~~~').format(IoU_lo_thresh, IoU_hi_thresh))\n print('{:.1f}'.format(100 * ap_default))\n for cls_ind, cls in enumerate(self.classes):\n if cls == '__background__':\n continue\n # minus 1 because of __background__\n precision = coco_eval.eval['precision'][ind_lo:(ind_hi + 1), :, cls_ind - 1, 0, 2]\n ap = np.mean(precision[precision > -1])\n print('{}: {:.1f}'.format(cls, 100 * ap))\n\n print('~~~~ Summary metrics ~~~~')\n coco_eval.summarize()\n\n def _do_detection_eval(self, res_file, output_dir):\n ann_type = 'bbox'\n coco_dt = self._COCO.loadRes(res_file)\n coco_eval = COCOeval(self._COCO, coco_dt)\n coco_eval.params.useSegm = (ann_type == 'segm')\n coco_eval.evaluate()\n coco_eval.accumulate()\n self._print_detection_eval_metrics(coco_eval)\n eval_file = osp.join(output_dir, 'detection_results.pkl')\n with open(eval_file, 'wb') as fid:\n pickle.dump(coco_eval, fid, pickle.HIGHEST_PROTOCOL)\n print('Wrote COCO eval results to: {}'.format(eval_file))\n\n def _coco_results_one_category(self, boxes, cat_id):\n results = []\n for im_ind, index in enumerate(self.image_index):\n dets = boxes[im_ind].astype(np.float)\n if dets == []:\n continue\n scores = dets[:, -1]\n xs = dets[:, 0]\n ys = dets[:, 1]\n ws = dets[:, 2] - xs + 1\n hs = dets[:, 3] - ys + 1\n results.extend(\n [{'image_id': index,\n 'category_id': cat_id,\n 'bbox': [xs[k], ys[k], ws[k], hs[k]],\n 'score': scores[k]} for k in range(dets.shape[0])])\n return results\n\n def _write_coco_results_file(self, all_boxes, res_file):\n # [{\"image_id\": 42,\n # \"category_id\": 18,\n # \"bbox\": [258.15,41.29,348.26,243.78],\n # \"score\": 0.236}, ...]\n results = []\n for cls_ind, cls in enumerate(self.classes):\n if cls == '__background__':\n continue\n print('Collecting {} results ({:d}/{:d})'.format(cls, cls_ind,\n self.num_classes - 1))\n coco_cat_id = self._class_to_coco_cat_id[cls]\n results.extend(self._coco_results_one_category(all_boxes[cls_ind],\n coco_cat_id))\n print('Writing results json to {}'.format(res_file))\n with open(res_file, 'w') as fid:\n json.dump(results, fid)\n\n def evaluate_detections(self, all_boxes, output_dir):\n res_file = osp.join(output_dir, ('detections_' +\n self._image_set +\n self._year +\n '_results'))\n if self.config['use_salt']:\n res_file += '_{}'.format(str(uuid.uuid4()))\n res_file += '.json'\n self._write_coco_results_file(all_boxes, res_file)\n # Only do evaluation on non-test sets\n if self._image_set.find('test') == -1:\n self._do_detection_eval(res_file, output_dir)\n # Optionally cleanup results json file\n if self.config['cleanup']:\n os.remove(res_file)\n\n def competition_mode(self, on):\n if on:\n self.config['use_salt'] = False\n self.config['cleanup'] = False\n else:\n self.config['use_salt'] = True\n self.config['cleanup'] = True"
},
{
"path": "lib/datasets/ds_utils.py",
"content": "# --------------------------------------------------------\n# Fast/er R-CNN\n# Licensed under The MIT License [see LICENSE for details]\n# Written by Ross Girshick\n# --------------------------------------------------------\nfrom __future__ import absolute_import\nfrom __future__ import division\nfrom __future__ import print_function\n\nimport numpy as np\n\n\ndef unique_boxes(boxes, scale=1.0):\n \"\"\"Return indices of unique boxes.\"\"\"\n v = np.array([1, 1e3, 1e6, 1e9])\n hashes = np.round(boxes * scale).dot(v)\n _, index = np.unique(hashes, return_index=True)\n return np.sort(index)\n\n\ndef xywh_to_xyxy(boxes):\n \"\"\"Convert [x y w h] box format to [x1 y1 x2 y2] format.\"\"\"\n return np.hstack((boxes[:, 0:2], boxes[:, 0:2] + boxes[:, 2:4] - 1))\n\n\ndef xyxy_to_xywh(boxes):\n \"\"\"Convert [x1 y1 x2 y2] box format to [x y w h] format.\"\"\"\n return np.hstack((boxes[:, 0:2], boxes[:, 2:4] - boxes[:, 0:2] + 1))\n\n\ndef validate_boxes(boxes, width=0, height=0):\n \"\"\"Check that a set of boxes are valid.\"\"\"\n x1 = boxes[:, 0]\n y1 = boxes[:, 1]\n x2 = boxes[:, 2]\n y2 = boxes[:, 3]\n assert (x1 >= 0).all()\n assert (y1 >= 0).all()\n assert (x2 >= x1).all()\n assert (y2 >= y1).all()\n assert (x2 < width).all()\n assert (y2 < height).all()\n\n\ndef filter_small_boxes(boxes, min_size):\n w = boxes[:, 2] - boxes[:, 0]\n h = boxes[:, 3] - boxes[:, 1]\n keep = np.where((w >= min_size) & (h > min_size))[0]\n return keep\n"
},
{
"path": "lib/datasets/factory.py",
"content": "# --------------------------------------------------------\n# Fast R-CNN\n# Copyright (c) 2015 Microsoft\n# Licensed under The MIT License [see LICENSE for details]\n# Written by Ross Girshick\n# --------------------------------------------------------\n\n\"\"\"Factory method for easily getting imdbs by name.\"\"\"\nfrom __future__ import absolute_import\nfrom __future__ import division\nfrom __future__ import print_function\n\n__sets = {}\n\nfrom datasets.coco import coco\nfrom datasets.pascal_voc import pascal_voc\n\n# # Set up voc__\nfor year in ['2007', '2012']:\n for split in ['train', 'val', 'trainval', 'test','shots', 'train_first_split', 'train_second_split', 'train_third_split']:\n name = 'voc_{}_{}'.format(year, split)\n __sets[name] = (lambda split=split, year=year: pascal_voc(split, year))\n\n\nfor year in ['2014']:\n for split in ['train', 'val', 'minival', 'valminusminival', 'trainval']:\n name = 'coco_{}_{}'.format(year, split)\n __sets[name] = (lambda split=split, year=year: coco(split, year))\n\nfor year in ['2017']:\n for split in ['train', 'val']:\n name = 'coco_{}_{}'.format(year, split)\n __sets[name] = (lambda split=split, year=year: coco(split, year))\n\ndef get_imdb(name):\n \"\"\"Get an imdb (image database) by name.\"\"\"\n if name not in __sets:\n raise KeyError('Unknown dataset: {}'.format(name))\n return __sets[name]()\n\n\ndef list_imdbs():\n \"\"\"List all registered imdbs.\"\"\"\n return list(__sets.keys())\n"
},
{
"path": "lib/datasets/imdb.py",
"content": "# --------------------------------------------------------\n# Fast R-CNN\n# Copyright (c) 2015 Microsoft\n# Licensed under The MIT License [see LICENSE for details]\n# Written by Ross Girshick and Xinlei Chen\n# --------------------------------------------------------\nfrom __future__ import absolute_import\nfrom __future__ import division\nfrom __future__ import print_function\n\nimport os\nimport os.path as osp\nimport PIL\nfrom model.utils.cython_bbox import bbox_overlaps\nimport numpy as np\nimport scipy.sparse\nfrom model.utils.config import cfg\nimport pdb\n\nROOT_DIR = osp.join(osp.dirname(__file__), '..', '..')\n\nclass imdb(object):\n \"\"\"Image database.\"\"\"\n\n def __init__(self, name, classes=None):\n self._name = name\n self._num_classes = 0\n if not classes:\n self._classes = []\n else:\n self._classes = classes\n self._image_index = []\n self._obj_proposer = 'gt'\n self._roidb = None\n self._roidb_handler = self.default_roidb\n # Use this dict for storing dataset specific config options\n self.config = {}\n\n @property\n def name(self):\n return self._name\n\n @property\n def num_classes(self):\n return len(self._classes)\n\n def set_classes(self,classes):\n self._classes = classes\n\n @property\n def classes(self):\n return self._classes\n\n @property\n def image_index(self):\n return self._image_index\n\n @property\n def roidb_handler(self):\n return self._roidb_handler\n\n @roidb_handler.setter\n def roidb_handler(self, val):\n self._roidb_handler = val\n\n def set_proposal_method(self, method):\n method = eval('self.' + method + '_roidb')\n self.roidb_handler = method\n\n @property\n def roidb(self):\n # A roidb is a list of dictionaries, each with the following keys:\n # boxes\n # gt_overlaps\n # gt_classes\n # flipped\n if self._roidb is not None:\n return self._roidb\n self._roidb = self.roidb_handler()\n return self._roidb\n\n def set_roidb(self,roidb):\n self._roidb = roidb\n\n\n\n @property\n def cache_path(self):\n cache_path = osp.abspath(osp.join(cfg.DATA_DIR, 'cache'))\n if not os.path.exists(cache_path):\n os.makedirs(cache_path)\n return cache_path\n\n @property\n def num_images(self):\n return len(self.image_index)\n\n def image_path_at(self, i):\n raise NotImplementedError\n\n def image_id_at(self, i):\n raise NotImplementedError\n\n def default_roidb(self):\n raise NotImplementedError\n\n def evaluate_detections(self, all_boxes, output_dir=None):\n \"\"\"\n all_boxes is a list of length number-of-classes.\n Each list element is a list of length number-of-images.\n Each of those list elements is either an empty list []\n or a numpy array of detection.\n\n all_boxes[class][image] = [] or np.array of shape #dets x 5\n \"\"\"\n raise NotImplementedError\n\n def _get_widths(self):\n return [PIL.Image.open(self.image_path_at(i)).size[0]\n for i in range(self.num_images)]\n\n def append_flipped_images(self):\n num_images = self.num_images\n widths = self._get_widths()\n for i in range(num_images):\n boxes = self.roidb[i]['boxes'].copy()\n oldx1 = boxes[:, 0].copy()\n oldx2 = boxes[:, 2].copy()\n boxes[:, 0] = widths[i] - oldx2 - 1\n boxes[:, 2] = widths[i] - oldx1 - 1\n assert (boxes[:, 2] >= boxes[:, 0]).all()\n\n\n entry = {'boxes': boxes,\n 'gt_overlaps': self.roidb[i]['gt_overlaps'],\n 'gt_classes': self.roidb[i]['gt_classes'],\n 'flipped': True}\n self.roidb.append(entry)\n self._image_index = self._image_index * 2\n\n def evaluate_recall(self, candidate_boxes=None, thresholds=None,\n area='all', limit=None):\n \"\"\"Evaluate detection proposal recall metrics.\n\n Returns:\n results: dictionary of results with keys\n 'ar': average recall\n 'recalls': vector recalls at each IoU overlap threshold\n 'thresholds': vector of IoU overlap thresholds\n 'gt_overlaps': vector of all ground-truth overlaps\n \"\"\"\n # Record max overlap value for each gt box\n # Return vector of overlap values\n areas = {'all': 0, 'small': 1, 'medium': 2, 'large': 3,\n '96-128': 4, '128-256': 5, '256-512': 6, '512-inf': 7}\n area_ranges = [[0 ** 2, 1e5 ** 2], # all\n [0 ** 2, 32 ** 2], # small\n [32 ** 2, 96 ** 2], # medium\n [96 ** 2, 1e5 ** 2], # large\n [96 ** 2, 128 ** 2], # 96-128\n [128 ** 2, 256 ** 2], # 128-256\n [256 ** 2, 512 ** 2], # 256-512\n [512 ** 2, 1e5 ** 2], # 512-inf\n ]\n assert area in areas, 'unknown area range: {}'.format(area)\n area_range = area_ranges[areas[area]]\n gt_overlaps = np.zeros(0)\n num_pos = 0\n for i in range(self.num_images):\n # Checking for max_overlaps == 1 avoids including crowd annotations\n # (...pretty hacking :/)\n max_gt_overlaps = self.roidb[i]['gt_overlaps'].toarray().max(axis=1)\n gt_inds = np.where((self.roidb[i]['gt_classes'] > 0) &\n (max_gt_overlaps == 1))[0]\n gt_boxes = self.roidb[i]['boxes'][gt_inds, :]\n gt_areas = self.roidb[i]['seg_areas'][gt_inds]\n valid_gt_inds = np.where((gt_areas >= area_range[0]) &\n (gt_areas <= area_range[1]))[0]\n gt_boxes = gt_boxes[valid_gt_inds, :]\n num_pos += len(valid_gt_inds)\n\n if candidate_boxes is None:\n # If candidate_boxes is not supplied, the default is to use the\n # non-ground-truth boxes from this roidb\n non_gt_inds = np.where(self.roidb[i]['gt_classes'] == 0)[0]\n boxes = self.roidb[i]['boxes'][non_gt_inds, :]\n else:\n boxes = candidate_boxes[i]\n if boxes.shape[0] == 0:\n continue\n if limit is not None and boxes.shape[0] > limit:\n boxes = boxes[:limit, :]\n\n overlaps = bbox_overlaps(boxes.astype(np.float),\n gt_boxes.astype(np.float))\n\n _gt_overlaps = np.zeros((gt_boxes.shape[0]))\n for j in range(gt_boxes.shape[0]):\n # find which proposal box maximally covers each gt box\n argmax_overlaps = overlaps.argmax(axis=0)\n # and get the iou amount of coverage for each gt box\n max_overlaps = overlaps.max(axis=0)\n # find which gt box is 'best' covered (i.e. 'best' = most iou)\n gt_ind = max_overlaps.argmax()\n gt_ovr = max_overlaps.max()\n assert (gt_ovr >= 0)\n # find the proposal box that covers the best covered gt box\n box_ind = argmax_overlaps[gt_ind]\n # record the iou coverage of this gt box\n _gt_overlaps[j] = overlaps[box_ind, gt_ind]\n assert (_gt_overlaps[j] == gt_ovr)\n # mark the proposal box and the gt box as used\n overlaps[box_ind, :] = -1\n overlaps[:, gt_ind] = -1\n # append recorded iou coverage level\n gt_overlaps = np.hstack((gt_overlaps, _gt_overlaps))\n\n gt_overlaps = np.sort(gt_overlaps)\n if thresholds is None:\n step = 0.05\n thresholds = np.arange(0.5, 0.95 + 1e-5, step)\n recalls = np.zeros_like(thresholds)\n # compute recall for each iou threshold\n for i, t in enumerate(thresholds):\n recalls[i] = (gt_overlaps >= t).sum() / float(num_pos)\n # ar = 2 * np.trapz(recalls, thresholds)\n ar = recalls.mean()\n return {'ar': ar, 'recalls': recalls, 'thresholds': thresholds,\n 'gt_overlaps': gt_overlaps}\n\n def create_roidb_from_box_list(self, box_list, gt_roidb):\n assert len(box_list) == self.num_images, \\\n 'Number of boxes must match number of ground-truth images'\n roidb = []\n for i in range(self.num_images):\n boxes = box_list[i]\n num_boxes = boxes.shape[0]\n overlaps = np.zeros((num_boxes, self.num_classes), dtype=np.float32)\n\n if gt_roidb is not None and gt_roidb[i]['boxes'].size > 0:\n gt_boxes = gt_roidb[i]['boxes']\n gt_classes = gt_roidb[i]['gt_classes']\n gt_overlaps = bbox_overlaps(boxes.astype(np.float),\n gt_boxes.astype(np.float))\n argmaxes = gt_overlaps.argmax(axis=1)\n maxes = gt_overlaps.max(axis=1)\n I = np.where(maxes > 0)[0]\n overlaps[I, gt_classes[argmaxes[I]]] = maxes[I]\n\n overlaps = scipy.sparse.csr_matrix(overlaps)\n roidb.append({\n 'boxes': boxes,\n 'gt_classes': np.zeros((num_boxes,), dtype=np.int32),\n 'gt_overlaps': overlaps,\n 'flipped': False,\n 'seg_areas': np.zeros((num_boxes,), dtype=np.float32),\n })\n return roidb\n\n @staticmethod\n def merge_roidbs(a, b):\n assert len(a) == len(b)\n for i in range(len(a)):\n a[i]['boxes'] = np.vstack((a[i]['boxes'], b[i]['boxes']))\n a[i]['gt_classes'] = np.hstack((a[i]['gt_classes'],\n b[i]['gt_classes']))\n a[i]['gt_overlaps'] = scipy.sparse.vstack([a[i]['gt_overlaps'],\n b[i]['gt_overlaps']])\n a[i]['seg_areas'] = np.hstack((a[i]['seg_areas'],\n b[i]['seg_areas']))\n return a\n\n def competition_mode(self, on):\n \"\"\"Turn competition mode on or off.\"\"\"\n pass\n"
},
{
"path": "lib/datasets/metadata.py",
"content": "# --------------------------------------------------------\n# Pytorch Meta R-CNN\n# Written by Anny Xu, Xiaopeng Yan, based on the code from Jianwei Yang\n# --------------------------------------------------------\nimport os\nimport os.path\nimport sys\nimport torch.utils.data as data\nimport cv2\nimport torch\nimport random\nimport numpy as np\nif sys.version_info[0] == 2:\n import xml.etree.cElementTree as ET\nelse:\n import xml.etree.ElementTree as ET\nfrom model.utils.config import cfg\nimport collections\n\nclass MetaDataset(data.Dataset):\n\n \"\"\"Meta Dataset\n Arguments:\n root (string): filepath to VOCdevkit folder.\n image_set (string): imageset to use (eg. 'train', 'val')\n metaclass(string): the class name\n img_size(int) : the PRN network input size\n shot(int): the number of instances\n shuffle(bool)\n \"\"\"\n\n def __init__(self, root, image_sets, metaclass, img_size, shots=1, shuffle=False, phase=1):\n self.root = root\n self.image_set = image_sets\n self.img_size = img_size\n self.metaclass = metaclass\n self.shots = shots\n if phase == 2:\n self.shots = shots * 3\n self.shuffle=shuffle\n self._annopath = os.path.join('%s', 'Annotations', '%s.xml')\n self._imgpath = os.path.join('%s', 'JPEGImages', '%s.jpg')\n self.shot_path = open(os.path.join(self.root, 'VOC2007', 'ImageSets/Main/shots.txt'), 'w') # the default saved path\n self.ids = list()\n for (year, name) in image_sets:\n self._year = year\n rootpath = os.path.join(self.root, 'VOC' + year)\n for line in open(os.path.join(rootpath, 'ImageSets', 'Main', name + '.txt')):\n self.ids.append((rootpath, line.strip()))\n\n class_to_idx = dict(zip(self.metaclass, range(len(self.metaclass)))) # class to index mapping\n\n self.prndata = []\n self.prncls = []\n prn_image, prn_mask = self.get_prndata()\n for i in range(shots):\n cls = []\n data = []\n for n, key in enumerate(list(prn_image.keys())):\n img = torch.from_numpy(np.array(prn_image[key][i]))\n img = img.unsqueeze(0)\n mask = torch.from_numpy(np.array(prn_mask[key][i]))\n mask = mask.unsqueeze(0)\n mask = mask.unsqueeze(3)\n imgmask = torch.cat([img, mask], dim=3)\n data.append(imgmask.permute(0, 3, 1, 2).contiguous())\n cls.append(class_to_idx[key])\n self.prncls.append(cls)\n self.prndata.append(torch.cat(data,dim=0))\n\n def __getitem__(self, index):\n return self.prndata[index],self.prncls[index]\n\n def get_prndata(self):\n '''\n :return: the construct prn input data\n '''\n if self.shuffle:\n random.shuffle(self.ids)\n prn_image = collections.defaultdict(list)\n prn_mask = collections.defaultdict(list)\n classes = collections.defaultdict(int)\n for cls in self.metaclass:\n classes[cls] = 0\n for img_id in self.ids:\n target = ET.parse(self._annopath % img_id).getroot()\n img = cv2.imread(self._imgpath % img_id, cv2.IMREAD_COLOR)\n img = img[:, :, ::-1]\n img = img.astype(np.float32, copy=False)\n img -= cfg.PIXEL_MEANS\n height, width, _ = img.shape\n mask = np.zeros((self.img_size, self.img_size), dtype=np.float32)\n h, w, _ = img.shape\n y_ration = float(h) / self.img_size\n x_ration = float(w) / self.img_size\n img_resize = cv2.resize(img, (self.img_size, self.img_size), interpolation=cv2.INTER_LINEAR)\n for obj in target.iter('object'):\n difficult = int(obj.find('difficult').text) == 1\n if difficult:\n continue\n name = obj.find('name').text.strip()\n if name not in self.metaclass:\n continue\n if classes[name] >= self.shots:\n break\n classes[name] += 1\n bbox = obj.find('bndbox')\n pts = ['xmin', 'ymin', 'xmax', 'ymax']\n bndbox = []\n for i, pt in enumerate(pts):\n cur_pt = int(float(bbox.find(pt).text)) - 1\n if i % 2 == 0:\n cur_pt = int(cur_pt / x_ration)\n bndbox.append(cur_pt)\n elif i % 2 == 1:\n cur_pt = int(cur_pt / y_ration)\n bndbox.append(cur_pt)\n mask[bndbox[1]:bndbox[3], bndbox[0]:bndbox[2]] = 1\n prn_image[name].append(img_resize)\n prn_mask[name].append(mask)\n self.shot_path.write(str(img_id[1])+'\\n')\n break\n if len(classes)>0 and min(classes.values()) == self.shots:\n break\n self.shot_path.close()\n return prn_image, prn_mask\n\n def __len__(self):\n return len(self.prndata)\n"
},
{
"path": "lib/datasets/pascal_voc.py",
"content": "from __future__ import print_function\nfrom __future__ import absolute_import\n# --------------------------------------------------------\n# Fast R-CNN\n# Copyright (c) 2015 Microsoft\n# Licensed under The MIT License [see LICENSE for details]\n# Written by Ross Girshick\n# --------------------------------------------------------\n\nimport xml.dom.minidom as minidom\n\nimport os\n# import PIL\nimport numpy as np\nimport scipy.sparse\nimport subprocess\nimport math\nimport glob\nimport uuid\nimport scipy.io as sio\nimport xml.etree.ElementTree as ET\nimport pickle\nfrom .imdb import imdb\nfrom .imdb import ROOT_DIR\nfrom . import ds_utils\nfrom .voc_eval import voc_eval\nimport random\n# TODO: make fast_rcnn irrelevant\n# >>>> obsolete, because it depends on sth outside of this project\nfrom model.utils.config import cfg\n\ntry:\n xrange # Python 2\nexcept NameError:\n xrange = range # Python 3\n\n# <<<< obsolete\n\n\nclass pascal_voc(imdb):\n def __init__(self, image_set, year, devkit_path=None):\n imdb.__init__(self, 'voc_' + year + '_' + image_set)\n self._year = year\n self._image_set = image_set\n self._devkit_path = self._get_default_path() if devkit_path is None \\\n else devkit_path\n self._data_path = os.path.join(self._devkit_path, 'VOC' + self._year)\n #first split\n if cfg.TRAIN.META_TYPE == 1:\n self._classes = ['__background__'] + cfg.TRAIN.ALLCLASSES_FIRST\n #second split\n if cfg.TRAIN.META_TYPE == 2:\n self._classes = ['__background__'] + cfg.TRAIN.ALLCLASSES_SECOND\n #third split\n if cfg.TRAIN.META_TYPE == 3:\n self._classes = ['__background__'] + cfg.TRAIN.ALLCLASSES_THIRD\n\n self._class_to_ind = dict(zip(self.classes, xrange(self.num_classes)))\n self._image_ext = '.jpg'\n self._image_index = self._load_image_set_index()\n # Default to roidb handler\n # self._roidb_handler = self.selective_search_roidb\n self._roidb_handler = self.gt_roidb\n self._salt = str(uuid.uuid4())\n self._comp_id = 'comp4'\n\n # PASCAL specific config options\n self.config = {'cleanup': True,\n 'use_salt': True,\n 'use_diff': False,\n 'matlab_eval': False,\n 'rpn_file': None,\n 'min_size': 2}\n\n assert os.path.exists(self._devkit_path), \\\n 'VOCdevkit path does not exist: {}'.format(self._devkit_path)\n assert os.path.exists(self._data_path), \\\n 'Path does not exist: {}'.format(self._data_path)\n\n def image_path_at(self, i):\n \"\"\"\n Return the absolute path to image i in the image sequence.\n \"\"\"\n return self.image_path_from_index(self._image_index[i])\n\n def image_id_at(self, i):\n \"\"\"\n Return the absolute path to image i in the image sequence.\n \"\"\"\n return i\n\n def image_path_from_index(self, index):\n \"\"\"\n Construct an image path from the image's \"index\" identifier.\n \"\"\"\n # file_name = str(index).zfill(6)\n image_path = os.path.join(self._data_path, 'JPEGImages',\n index + self._image_ext)\n assert os.path.exists(image_path), \\\n 'Path does not exist: {}'.format(image_path)\n return image_path\n\n def _load_image_set_index(self):\n \"\"\"\n Load the indexes listed in this dataset's image set file.\n \"\"\"\n # Example path to image set file:\n # self._devkit_path + /VOCdevkit2007/VOC2007/ImageSets/Main/val.txt\n image_set_file = os.path.join(self._data_path, 'ImageSets', 'Main',\n self._image_set + '.txt')\n assert os.path.exists(image_set_file), \\\n 'Path does not exist: {}'.format(image_set_file)\n with open(image_set_file) as f:\n image_index = [x.strip() for x in f.readlines()]\n return image_index\n\n def _get_default_path(self):\n \"\"\"\n Return the default path where PASCAL VOC is expected to be installed.\n \"\"\"\n return os.path.join(cfg.DATA_DIR, 'VOCdevkit' + self._year)\n\n def gt_roidb(self):\n \"\"\"\n Return the database of ground-truth regions of interest.\n\n This function loads/saves from/to a cache file to speed up future calls.\n \"\"\"\n cache_file = os.path.join(self.cache_path, self.name + '_gt_roidb.pkl')\n if os.path.exists(cache_file):\n with open(cache_file, 'rb') as fid:\n roidb = pickle.load(fid)\n print('{} gt roidb loaded from {}'.format(self.name, cache_file))\n return roidb\n\n gt_roidb = [self._load_pascal_annotation(index)\n for index in self.image_index]\n with open(cache_file, 'wb') as fid:\n pickle.dump(gt_roidb, fid, pickle.HIGHEST_PROTOCOL)\n print('wrote gt roidb to {}'.format(cache_file))\n\n return gt_roidb\n\n def selective_search_roidb(self):\n \"\"\"\n Return the database of selective search regions of interest.\n Ground-truth ROIs are also included.\n\n This function loads/saves from/to a cache file to speed up future calls.\n \"\"\"\n cache_file = os.path.join(self.cache_path,\n self.name + '_selective_search_roidb.pkl')\n\n if os.path.exists(cache_file):\n with open(cache_file, 'rb') as fid:\n roidb = pickle.load(fid)\n print('{} ss roidb loaded from {}'.format(self.name, cache_file))\n return roidb\n\n if int(self._year) == 2007 or self._image_set != 'test':\n gt_roidb = self.gt_roidb()\n ss_roidb = self._load_selective_search_roidb(gt_roidb)\n roidb = imdb.merge_roidbs(gt_roidb, ss_roidb)\n else:\n roidb = self._load_selective_search_roidb(None)\n with open(cache_file, 'wb') as fid:\n pickle.dump(roidb, fid, pickle.HIGHEST_PROTOCOL)\n print('wrote ss roidb to {}'.format(cache_file))\n\n return roidb\n\n def rpn_roidb(self):\n if int(self._year) == 2007 or self._image_set != 'test':\n gt_roidb = self.gt_roidb()\n rpn_roidb = self._load_rpn_roidb(gt_roidb)\n roidb = imdb.merge_roidbs(gt_roidb, rpn_roidb)\n else:\n roidb = self._load_rpn_roidb(None)\n\n return roidb\n\n def _load_rpn_roidb(self, gt_roidb):\n filename = self.config['rpn_file']\n print('loading {}'.format(filename))\n assert os.path.exists(filename), \\\n 'rpn data not found at: {}'.format(filename)\n with open(filename, 'rb') as f:\n box_list = pickle.load(f)\n return self.create_roidb_from_box_list(box_list, gt_roidb)\n\n def _load_selective_search_roidb(self, gt_roidb):\n filename = os.path.abspath(os.path.join(cfg.DATA_DIR,\n 'selective_search_data',\n self.name + '.mat'))\n assert os.path.exists(filename), \\\n 'Selective search data not found at: {}'.format(filename)\n raw_data = sio.loadmat(filename)['boxes'].ravel()\n\n box_list = []\n for i in xrange(raw_data.shape[0]):\n boxes = raw_data[i][:, (1, 0, 3, 2)] - 1\n keep = ds_utils.unique_boxes(boxes)\n boxes = boxes[keep, :]\n keep = ds_utils.filter_small_boxes(boxes, self.config['min_size'])\n boxes = boxes[keep, :]\n box_list.append(boxes)\n\n return self.create_roidb_from_box_list(box_list, gt_roidb)\n\n def _load_pascal_annotation(self, index):\n \"\"\"\n Load image and bounding boxes info from XML file in the PASCAL VOC\n format.\n \"\"\"\n filename = os.path.join(self._data_path, 'Annotations', index + '.xml')\n tree = ET.parse(filename)\n objs = tree.findall('object')\n # if not self.config['use_diff']:\n # # Exclude the samples labeled as difficult\n # non_diff_objs = [\n # obj for obj in objs if int(obj.find('difficult').text) == 0]\n # # if len(non_diff_objs) != len(objs):\n # # print 'Removed {} difficult objects'.format(\n # # len(objs) - len(non_diff_objs))\n # objs = non_diff_objs\n num_objs = len(objs)\n\n boxes = np.zeros((num_objs, 4), dtype=np.uint16)\n gt_classes = np.zeros((num_objs), dtype=np.int32)\n overlaps = np.zeros((num_objs, self.num_classes), dtype=np.float32)\n # \"Seg\" area for pascal is just the box area\n seg_areas = np.zeros((num_objs), dtype=np.float32)\n ishards = np.zeros((num_objs), dtype=np.int32)\n\n # Load object bounding boxes into a data frame.\n for ix, obj in enumerate(objs):\n bbox = obj.find('bndbox')\n # Make pixel indexes 0-based\n x1 = float(bbox.find('xmin').text) - 1\n y1 = float(bbox.find('ymin').text) - 1\n x2 = float(bbox.find('xmax').text) - 1\n y2 = float(bbox.find('ymax').text) - 1\n\n diffc = obj.find('difficult')\n difficult = 0 if diffc == None else int(diffc.text)\n ishards[ix] = difficult\n if obj.find('name').text.lower().strip() not in self._classes:\n continue\n cls = self._class_to_ind[obj.find('name').text.lower().strip()]\n boxes[ix, :] = [x1, y1, x2, y2]\n gt_classes[ix] = cls\n overlaps[ix, cls] = 1.0\n seg_areas[ix] = (x2 - x1 + 1) * (y2 - y1 + 1)\n\n overlaps = scipy.sparse.csr_matrix(overlaps)\n\n return {'boxes': boxes,\n 'gt_classes': gt_classes,\n 'gt_ishard': ishards,\n 'gt_overlaps': overlaps,\n 'flipped': False,\n 'seg_areas': seg_areas}\n\n def _get_comp_id(self):\n comp_id = (self._comp_id + '_' + self._salt if self.config['use_salt']\n else self._comp_id)\n return comp_id\n\n def _get_voc_results_file_template(self):\n # VOCdevkit/results/VOC2007/Main/_det_test_aeroplane.txt\n filename = self._get_comp_id() + '_det_' + self._image_set + '_{:s}.txt'\n filedir = os.path.join(self._devkit_path, 'results', 'VOC' + self._year, 'Main')\n if not os.path.exists(filedir):\n os.makedirs(filedir)\n path = os.path.join(filedir, filename)\n return path\n\n def _write_voc_results_file(self, all_boxes):\n for cls_ind, cls in enumerate(self.classes):\n if cls == '__background__':\n continue\n print('Writing {} VOC results file'.format(cls))\n filename = self._get_voc_results_file_template().format(cls)\n with open(filename, 'wt') as f:\n for im_ind, index in enumerate(self.image_index):\n dets = all_boxes[cls_ind][im_ind]\n if dets == []:\n continue\n # the VOCdevkit expects 1-based indices\n for k in xrange(dets.shape[0]):\n f.write('{:s} {:.3f} {:.1f} {:.1f} {:.1f} {:.1f}\\n'.\n format(index, dets[k, -1],\n dets[k, 0] + 1, dets[k, 1] + 1,\n dets[k, 2] + 1, dets[k, 3] + 1))\n\n def _do_python_eval(self, output_dir='output', **kwargs):\n annopath = os.path.join(\n self._devkit_path,\n 'VOC' + self._year,\n 'Annotations',\n '{:s}.xml')\n imagesetfile = os.path.join(\n self._devkit_path,\n 'VOC' + self._year,\n 'ImageSets',\n 'Main',\n self._image_set + '.txt')\n cachedir = os.path.join(self._devkit_path, 'annotations_cache')\n aps = []\n import time,csv\n now = time.strftime(\"%Y-%m-%d-%H-%M-%S\")\n ############################### changed by xan 2019/1/31 begin################################\n save_dir = 'results'\n if not os.path.exists(save_dir):\n os.mkdir(save_dir)\n path = str(now) + '_'\n for k in kwargs:\n if k in ('checksession','checkepoch','checkpoint','meta_test','shots'):\n path = path + k + '_' + str(kwargs[k])\n csvfile = open(os.path.join(save_dir, path + '.csv'), 'w')\n ############################## end ###########################################################\n writer = csv.writer(csvfile)\n\n # The PASCAL VOC metric changed in 2010\n use_07_metric = True if int(self._year) < 2010 else False\n print('VOC07 metric? ' + ('Yes' if use_07_metric else 'No'))\n if not os.path.isdir(output_dir):\n os.mkdir(output_dir)\n cls_names = []\n ap_values = []\n for i, cls in enumerate(self._classes):\n if cls == '__background__':\n continue\n filename = self._get_voc_results_file_template().format(cls)\n rec, prec, ap = voc_eval(\n filename, annopath, imagesetfile, cls, cachedir, ovthresh=0.5,\n use_07_metric=use_07_metric)\n aps += [ap]\n print('AP for {} = {:.3f}'.format(cls, ap))\n cls_names.append(cls)\n ap_values.append((\"%.1f\" % (ap*100)))\n if i == 15:\n cls_names.append('mean')\n tmp = np.mean(aps)*100\n ap_values.append((\"%.1f\" % tmp))\n if i == 20:\n cls_names.append('mean')\n tmp = np.mean(aps[-5:])*100\n ap_values.append((\"%.1f\" % tmp))\n with open(os.path.join(output_dir, cls + '_pr.pkl'), 'wb') as f:\n pickle.dump({'rec': rec, 'prec': prec, 'ap': ap}, f)\n print('Mean AP = {:.4f}'.format(np.mean(aps)))\n writer.writerow(cls_names)\n writer.writerow(ap_values)\n csvfile.close()\n print('~~~~~~~~')\n print('Results:')\n for ap in aps:\n print('{:.3f}'.format(ap))\n print('{:.3f}'.format(np.mean(aps)))\n print('~~~~~~~~')\n print('')\n print('--------------------------------------------------------------')\n print('Results computed with the **unofficial** Python eval code.')\n print('Results should be very close to the official MATLAB eval code.')\n print('Recompute with `./tools/reval.py --matlab ...` for your paper.')\n print('-- Thanks, The Management')\n print('--------------------------------------------------------------')\n\n def _do_matlab_eval(self, output_dir='output'):\n print('-----------------------------------------------------')\n print('Computing results with the official MATLAB eval code.')\n print('-----------------------------------------------------')\n path = os.path.join(cfg.ROOT_DIR, 'lib', 'datasets',\n 'VOCdevkit-matlab-wrapper')\n cmd = 'cd {} && '.format(path)\n cmd += '{:s} -nodisplay -nodesktop '.format(cfg.MATLAB)\n cmd += '-r \"dbstop if error; '\n cmd += 'voc_eval(\\'{:s}\\',\\'{:s}\\',\\'{:s}\\',\\'{:s}\\'); quit;\"' \\\n .format(self._devkit_path, self._get_comp_id(),\n self._image_set, output_dir)\n print('Running:\\n{}'.format(cmd))\n status = subprocess.call(cmd, shell=True)\n\n def evaluate_detections(self, all_boxes, output_dir, **kwargs):\n self._write_voc_results_file(all_boxes)\n self._do_python_eval(output_dir, **kwargs)\n if self.config['matlab_eval']:\n self._do_matlab_eval(output_dir)\n if self.config['cleanup']:\n for cls in self._classes:\n if cls == '__background__':\n continue\n filename = self._get_voc_results_file_template().format(cls)\n os.remove(filename)\n\n def competition_mode(self, on):\n if on:\n self.config['use_salt'] = False\n self.config['cleanup'] = False\n else:\n self.config['use_salt'] = True\n self.config['cleanup'] = True\n\n\nif __name__ == '__main__':\n d = pascal_voc('trainval', '2007')\n res = d.roidb\n from IPython import embed;\n\n embed()\n"
},
{
"path": "lib/datasets/pascal_voc_rbg.py",
"content": "# --------------------------------------------------------\n# Fast R-CNN\n# Copyright (c) 2015 Microsoft\n# Licensed under The MIT License [see LICENSE for details]\n# Written by Ross Girshick and Xinlei Chen\n# --------------------------------------------------------\nfrom __future__ import absolute_import\nfrom __future__ import division\nfrom __future__ import print_function\n\nimport os\nfrom datasets.imdb import imdb\nimport scipy.sparse\nimport pickle\nimport subprocess\nimport uuid\nfrom .voc_eval import voc_eval\nfrom model.utils.config import cfg\nimport pdb\nimport os\nimport os.path\nimport sys\nimport torch.utils.data as data\nimport cv2\nimport numpy as np\nif sys.version_info[0] == 2:\n import xml.etree.cElementTree as ET\nelse:\n import xml.etree.ElementTree as ET\n\nclass VOCDetection(data.Dataset):\n \"\"\"VOC Detection Dataset Object\n\n input is image, target is annotation\n\n Arguments:\n root (string): filepath to VOCdevkit folder.\n image_set (string): imageset to use (eg. 'train', 'val', 'test')\n transform (callable, optional): transformation to perform on the\n input image\n target_transform (callable, optional): transformation to perform on the\n target `annotation`\n (eg: take in caption string, return tensor of word indices)\n dataset_name (string, optional): which dataset to load\n (default: 'VOC2007')\n \"\"\"\n\n def __init__(self, root, image_sets, img_size, preproc=None, target_transform=None,\n dataset_name='VOC0712'):\n self.root = root\n self.image_set = image_sets\n self.img_size = img_size\n self.preproc = preproc\n self.target_transform = target_transform\n self.name = dataset_name\n self._annopath = os.path.join('%s', 'Annotations', '%s.xml')\n self._imgpath = os.path.join('%s', 'JPEGImages', '%s.jpg')\n self.ids = list()\n for (year, name) in image_sets:\n self._year = year\n rootpath = os.path.join(self.root, 'VOC' + year)\n for line in open(os.path.join(rootpath, 'ImageSets', 'Main', name + '.txt')):\n self.ids.append((rootpath, line.strip()))\n\n def __getitem__(self, index):\n img_id = self.ids[index]\n target = ET.parse(self._annopath % img_id).getroot()\n img = cv2.imread(self._imgpath % img_id, cv2.IMREAD_COLOR)\n height, width, _ = img.shape\n mask = np.zeros((self.img_size, self.img_size), dtype=np.uint8)\n h, w, _ = img.shape\n y_ration = float(h) / self.img_size\n x_ration = float(w) / self.img_size\n img_resize = cv2.resize(img, (self.img_size, self.img_size))\n labels = []\n for obj in target.iter('object'):\n difficult = int(obj.find('difficult').text) == 1\n if difficult:\n continue\n name = obj.find('name').text.strip()\n labels.append(name)\n bbox = obj.find('bndbox')\n pts = ['xmin', 'ymin', 'xmax', 'ymax']\n bndbox = []\n\n for i, pt in enumerate(pts):\n cur_pt = int(float(bbox.find(pt).text)) - 1\n if i % 2 == 0:\n cur_pt = int(cur_pt / x_ration)\n bndbox.append(cur_pt)\n elif i % 2 == 1:\n cur_pt = int(cur_pt / y_ration)\n bndbox.append(cur_pt)\n mask[bndbox[0]:bndbox[2], bndbox[1]:bndbox[3]] = 1\n return img_resize, mask, labels\n\n def __len__(self):\n return len(self.ids)\n\n\nclass pascal_voc(imdb):\n def __init__(self, image_set, year, devkit_path=None):\n imdb.__init__(self, 'voc_' + year + '_' + image_set)\n self._year = year\n self._image_set = image_set\n self._devkit_path = self._get_default_path() if devkit_path is None \\\n else devkit_path\n\n \n self._data_path = os.path.join(self._devkit_path, 'VOC' + self._year)\n self._classes = ('__background__', # always index 0\n 'aeroplane', 'bicycle', 'bird', 'boat',\n 'bottle', 'bus', 'car', 'cat', 'chair',\n 'cow', 'diningtable', 'dog', 'horse',\n 'motorbike', 'person', 'pottedplant',\n 'sheep', 'sofa', 'train', 'tvmonitor')\n self._class_to_ind = dict(list(zip(self.classes, list(range(self.num_classes)))))\n self._image_ext = '.jpg'\n self._image_index = self._load_image_set_index()\n # Default to roidb handler\n self._roidb_handler = self.gt_roidb\n self._salt = str(uuid.uuid4())\n self._comp_id = 'comp4'\n\n # PASCAL specific config options\n self.config = {'cleanup': True,\n 'use_salt': True,\n 'use_diff': False,\n 'matlab_eval': False,\n 'rpn_file': None}\n\n assert os.path.exists(self._devkit_path), \\\n 'VOCdevkit path does not exist: {}'.format(self._devkit_path)\n assert os.path.exists(self._data_path), \\\n 'Path does not exist: {}'.format(self._data_path)\n\n def image_path_at(self, i):\n \"\"\"\n Return the absolute path to image i in the image sequence.\n \"\"\"\n return self.image_path_from_index(self._image_index[i])\n\n def image_path_from_index(self, index):\n \"\"\"\n Construct an image path from the image's \"index\" identifier.\n \"\"\"\n image_path = os.path.join(self._data_path, 'JPEGImages',\n index + self._image_ext)\n assert os.path.exists(image_path), \\\n 'Path does not exist: {}'.format(image_path)\n return image_path\n\n def _load_image_set_index(self):\n \"\"\"\n Load the indexes listed in this dataset's image set file.\n \"\"\"\n # Example path to image set file:\n # self._devkit_path + /VOCdevkit2007/VOC2007/ImageSets/Main/val.txt\n image_set_file = os.path.join(self._data_path, 'ImageSets', 'Main',\n self._image_set + '.txt')\n \n assert os.path.exists(image_set_file), \\\n 'Path does not exist: {}'.format(image_set_file)\n with open(image_set_file) as f:\n image_index = [x.strip() for x in f.readlines()]\n return image_index\n\n def _get_default_path(self):\n \"\"\"\n Return the default path where PASCAL VOC is expected to be installed.\n \"\"\"\n return os.path.join(cfg.DATA_DIR, 'VOCdevkit' + self._year)\n\n def gt_roidb(self):\n \"\"\"\n Return the database of ground-truth regions of interest.\n\n This function loads/saves from/to a cache file to speed up future calls.\n \"\"\"\n cache_file = os.path.join(self.cache_path, self.name + '_gt_roidb.pkl')\n if os.path.exists(cache_file):\n with open(cache_file, 'rb') as fid:\n try:\n roidb = pickle.load(fid)\n except:\n roidb = pickle.load(fid, encoding='bytes')\n print('{} gt roidb loaded from {}'.format(self.name, cache_file))\n return roidb\n\n gt_roidb = [self._load_pascal_annotation(index)\n for index in self.image_index]\n with open(cache_file, 'wb') as fid:\n pickle.dump(gt_roidb, fid, pickle.HIGHEST_PROTOCOL)\n print('wrote gt roidb to {}'.format(cache_file))\n\n return gt_roidb\n\n def rpn_roidb(self):\n if int(self._year) == 2007 or self._image_set != 'test':\n gt_roidb = self.gt_roidb()\n rpn_roidb = self._load_rpn_roidb(gt_roidb)\n roidb = imdb.merge_roidbs(gt_roidb, rpn_roidb)\n else:\n roidb = self._load_rpn_roidb(None)\n\n return roidb\n\n def _load_rpn_roidb(self, gt_roidb):\n filename = self.config['rpn_file']\n print('loading {}'.format(filename))\n assert os.path.exists(filename), \\\n 'rpn data not found at: {}'.format(filename)\n with open(filename, 'rb') as f:\n box_list = pickle.load(f)\n return self.create_roidb_from_box_list(box_list, gt_roidb)\n\n def _load_pascal_annotation(self, index):\n \"\"\"\n Load image and bounding boxes info from XML file in the PASCAL VOC\n format.\n \"\"\"\n filename = os.path.join(self._data_path, 'Annotations', index + '.xml')\n tree = ET.parse(filename)\n objs = tree.findall('object')\n if not self.config['use_diff']:\n # Exclude the samples labeled as difficult\n non_diff_objs = [\n obj for obj in objs if int(obj.find('difficult').text) == 0]\n # if len(non_diff_objs) != len(objs):\n # print 'Removed {} difficult objects'.format(\n # len(objs) - len(non_diff_objs))\n objs = non_diff_objs\n num_objs = len(objs)\n\n boxes = np.zeros((num_objs, 4), dtype=np.uint16)\n gt_classes = np.zeros((num_objs), dtype=np.int32)\n overlaps = np.zeros((num_objs, self.num_classes), dtype=np.float32)\n # \"Seg\" area for pascal is just the box area\n seg_areas = np.zeros((num_objs), dtype=np.float32)\n\n # Load object bounding boxes into a data frame.\n for ix, obj in enumerate(objs):\n bbox = obj.find('bndbox')\n # Make pixel indexes 0-based\n x1 = float(bbox.find('xmin').text) - 1\n y1 = float(bbox.find('ymin').text) - 1\n x2 = float(bbox.find('xmax').text) - 1\n y2 = float(bbox.find('ymax').text) - 1\n cls = self._class_to_ind[obj.find('name').text.lower().strip()]\n boxes[ix, :] = [x1, y1, x2, y2]\n gt_classes[ix] = cls\n overlaps[ix, cls] = 1.0\n seg_areas[ix] = (x2 - x1 + 1) * (y2 - y1 + 1)\n\n overlaps = scipy.sparse.csr_matrix(overlaps)\n\n return {'boxes': boxes,\n 'gt_classes': gt_classes,\n 'gt_overlaps': overlaps,\n 'flipped': False,\n 'seg_areas': seg_areas}\n\n def _get_comp_id(self):\n comp_id = (self._comp_id + '_' + self._salt if self.config['use_salt']\n else self._comp_id)\n return comp_id\n\n def _get_voc_results_file_template(self):\n # VOCdevkit/results/VOC2007/Main/_det_test_aeroplane.txt\n filename = self._get_comp_id() + '_det_' + self._image_set + '_{:s}.txt'\n path = os.path.join(\n self._devkit_path,\n 'results',\n 'VOC' + self._year,\n 'Main',\n filename)\n return path\n\n def _write_voc_results_file(self, all_boxes):\n for cls_ind, cls in enumerate(self.classes):\n if cls == '__background__':\n continue\n print('Writing {} VOC results file'.format(cls))\n filename = self._get_voc_results_file_template().format(cls)\n with open(filename, 'wt') as f:\n for im_ind, index in enumerate(self.image_index):\n dets = all_boxes[cls_ind][im_ind]\n if dets == []:\n continue\n # the VOCdevkit expects 1-based indices\n for k in range(dets.shape[0]):\n f.write('{:s} {:.3f} {:.1f} {:.1f} {:.1f} {:.1f}\\n'.\n format(index, dets[k, -1],\n dets[k, 0] + 1, dets[k, 1] + 1,\n dets[k, 2] + 1, dets[k, 3] + 1))\n\n def _do_python_eval(self, output_dir='output'):\n annopath = os.path.join(\n self._devkit_path,\n 'VOC' + self._year,\n 'Annotations',\n '{:s}.xml')\n imagesetfile = os.path.join(\n self._devkit_path,\n 'VOC' + self._year,\n 'ImageSets',\n 'Main',\n self._image_set + '.txt')\n cachedir = os.path.join(self._devkit_path, 'annotations_cache')\n aps = []\n # The PASCAL VOC metric changed in 2010\n use_07_metric = True if int(self._year) < 2010 else False\n print('VOC07 metric? ' + ('Yes' if use_07_metric else 'No'))\n if not os.path.isdir(output_dir):\n os.mkdir(output_dir)\n for i, cls in enumerate(self._classes):\n if cls == '__background__':\n continue\n filename = self._get_voc_results_file_template().format(cls)\n rec, prec, ap = voc_eval(\n filename, annopath, imagesetfile, cls, cachedir, ovthresh=0.5,\n use_07_metric=use_07_metric)\n aps += [ap]\n print(('AP for {} = {:.4f}'.format(cls, ap)))\n with open(os.path.join(output_dir, cls + '_pr.pkl'), 'wb') as f:\n pickle.dump({'rec': rec, 'prec': prec, 'ap': ap}, f)\n print(('Mean AP = {:.4f}'.format(np.mean(aps))))\n print('~~~~~~~~')\n print('Results:')\n for ap in aps:\n print(('{:.3f}'.format(ap)))\n print(('{:.3f}'.format(np.mean(aps))))\n print('~~~~~~~~')\n print('')\n print('--------------------------------------------------------------')\n print('Results computed with the **unofficial** Python eval code.')\n print('Results should be very close to the official MATLAB eval code.')\n print('Recompute with `./tools/reval.py --matlab ...` for your paper.')\n print('-- Thanks, The Management')\n print('--------------------------------------------------------------')\n\n def _do_matlab_eval(self, output_dir='output'):\n print('-----------------------------------------------------')\n print('Computing results with the official MATLAB eval code.')\n print('-----------------------------------------------------')\n path = os.path.join(cfg.ROOT_DIR, 'lib', 'datasets',\n 'VOCdevkit-matlab-wrapper')\n cmd = 'cd {} && '.format(path)\n cmd += '{:s} -nodisplay -nodesktop '.format(cfg.MATLAB)\n cmd += '-r \"dbstop if error; '\n cmd += 'voc_eval(\\'{:s}\\',\\'{:s}\\',\\'{:s}\\',\\'{:s}\\'); quit;\"' \\\n .format(self._devkit_path, self._get_comp_id(),\n self._image_set, output_dir)\n print(('Running:\\n{}'.format(cmd)))\n status = subprocess.call(cmd, shell=True)\n\n def evaluate_detections(self, all_boxes, output_dir):\n pdb.set_trace()\n self._write_voc_results_file(all_boxes)\n self._do_python_eval(output_dir)\n if self.config['matlab_eval']:\n self._do_matlab_eval(output_dir)\n if self.config['cleanup']:\n for cls in self._classes:\n if cls == '__background__':\n continue\n filename = self._get_voc_results_file_template().format(cls)\n os.remove(filename)\n\n def competition_mode(self, on):\n if on:\n self.config['use_salt'] = False\n self.config['cleanup'] = False\n else:\n self.config['use_salt'] = True\n self.config['cleanup'] = True\n\n\nif __name__ == '__main__':\n from datasets.pascal_voc import pascal_voc\n\n d = pascal_voc('trainval', '2007')\n res = d.roidb\n from IPython import embed;\n\n embed()\n"
},
{
"path": "lib/datasets/tools/compute_prior.py",
"content": "import numpy as np\nimport pickle\nimport os\nimport sys\n\nNUM_ATTR_REL = 200\ndef cout_w(prob, num=NUM_ATTR_REL,dim=1):\n prob_weight = prob[:, :num]\n sum_value = np.sum(prob_weight, keepdims=True, axis=dim) + 0.1\n prob_weight = prob_weight / np.repeat(sum_value, prob_weight.shape[dim], axis=dim)\n return prob_weight\n\ndef cp_kl(a, b):\n # compute kl diverse\n if np.sum(a) == 0 or np.sum(b) == 0:\n return 1\n sum_ = a * np.log(a / b)\n all_value = [x for x in sum_ if str(x) != 'nan' and str(x) != 'inf']\n kl = np.sum(all_value)\n return kl\n\ndef compute_js(attr_prob):\n cls_num = attr_prob.shape[0]\n similarity = np.zeros((cls_num, cls_num))\n similarity[0, 1:] = 1\n similarity[1:, 0] = 1\n for i in range(1, cls_num):\n if i % 50 == 0:\n print('had proccessed {} cls...\\n'.format(i))\n for j in range(1, cls_num):\n if i == j:\n similarity[i,j] = 0\n else:\n similarity[i,j] = 0.5 * (cp_kl(attr_prob[i, :], 0.5*(attr_prob[i, :] + attr_prob[j,:]))\n + cp_kl(attr_prob[j, :], 0.5*(attr_prob[i, :] + attr_prob[j, :])))\n return similarity\n\nif __name__=='__main__':\n data_path = '/data/VisualGenome/graph/'\n dim_ = 1000\n ## Compute attribute knowledge by JS-diversion\n graph_a = pickle.load(open(data_path + 'vg_attr_frequency_1000.pkl', 'rb'))\n\n ## You can get part of graph_a and match name with your datasets\n # We give an example of compute graph of VisualGenome with 1000 classes\n # first line of graph_a is background\n graph_a = cout_w(graph_a, num=len(graph_a))\n graph_a = compute_js(graph_a)\n graph_a = 1 - graph_a\n pickle.dump(graph_a, open(data_path + 'vg_graph_a.pkl', 'wb'))\n\n ## Compute relation knowledge\n graph_r = pickle.load(open(data_path + 'vg_pair_frequency_1000.pkl', 'rb'))\n ## You can get part of graph_a and match name with your datasets\n # We give an example of compute graph of VisualGenome with 1000 classes\n relation_matrix = np.zeros((dim_, dim_))\n relation_matrix = graph_r + graph_r.transpose()\n relation_matrix_row_sum = relation_matrix.sum(1)\n for i in range(dim_):\n relation_matrix[i, i] = relation_matrix_row_sum[i] + 1.\n prob_relation_matrix = np.zeros((dim_, dim_))\n for i in range(dim_):\n for j in range(dim_):\n prob_relation_matrix[i, j] = relation_matrix[i, j] / (\n np.sqrt(relation_matrix[i, i]) * np.sqrt(relation_matrix[j, j]))\n prob_relation_matrix_ba = np.zeros((dim_ + 1, dim_ + 1))\n prob_relation_matrix_ba[1:, 1:] = prob_relation_matrix\n print(prob_relation_matrix_ba.shape)\n pickle.dump(prob_relation_matrix_ba, open(data_path + 'vg_graph_r.pkl', 'wb'))\n"
},
{
"path": "lib/datasets/tools/mcg_munge.py",
"content": "from __future__ import print_function\nimport os\nimport sys\n\n\"\"\"Hacky tool to convert file system layout of MCG boxes downloaded from\nhttp://www.eecs.berkeley.edu/Research/Projects/CS/vision/grouping/mcg/\nso that it's consistent with those computed by Jan Hosang (see:\nhttp://www.mpi-inf.mpg.de/departments/computer-vision-and-multimodal-\n computing/research/object-recognition-and-scene-understanding/how-\n good-are-detection-proposals-really/)\n\nNB: Boxes from the MCG website are in (y1, x1, y2, x2) order.\nBoxes from Hosang et al. are in (x1, y1, x2, y2) order.\n\"\"\"\n\ndef munge(src_dir):\n # stored as: ./MCG-COCO-val2014-boxes/COCO_val2014_000000193401.mat\n # want: ./MCG/mat/COCO_val2014_0/COCO_val2014_000000141/COCO_val2014_000000141334.mat\n\n files = os.listdir(src_dir)\n for fn in files:\n base, ext = os.path.splitext(fn)\n # first 14 chars / first 22 chars / all chars + .mat\n # COCO_val2014_0/COCO_val2014_000000447/COCO_val2014_000000447991.mat\n first = base[:14]\n second = base[:22]\n dst_dir = os.path.join('MCG', 'mat', first, second)\n if not os.path.exists(dst_dir):\n os.makedirs(dst_dir)\n src = os.path.join(src_dir, fn)\n dst = os.path.join(dst_dir, fn)\n print('MV: {} -> {}'.format(src, dst))\n os.rename(src, dst)\n\nif __name__ == '__main__':\n # src_dir should look something like:\n # src_dir = 'MCG-COCO-val2014-boxes'\n src_dir = sys.argv[1]\n munge(src_dir)\n"
},
{
"path": "lib/datasets/voc_eval.py",
"content": "# --------------------------------------------------------\n# Fast/er R-CNN\n# Licensed under The MIT License [see LICENSE for details]\n# Written by Bharath Hariharan\n# --------------------------------------------------------\nfrom __future__ import absolute_import\nfrom __future__ import division\nfrom __future__ import print_function\n\nimport xml.etree.ElementTree as ET\nimport os\nimport pickle\nimport numpy as np\n\ndef parse_rec(filename):\n \"\"\" Parse a PASCAL VOC xml file \"\"\"\n tree = ET.parse(filename)\n objects = []\n for obj in tree.findall('object'):\n obj_struct = {}\n obj_struct['name'] = obj.find('name').text\n obj_struct['pose'] = obj.find('pose').text\n obj_struct['truncated'] = int(obj.find('truncated').text)\n obj_struct['difficult'] = int(obj.find('difficult').text)\n bbox = obj.find('bndbox')\n obj_struct['bbox'] = [int(bbox.find('xmin').text),\n int(bbox.find('ymin').text),\n int(bbox.find('xmax').text),\n int(bbox.find('ymax').text)]\n objects.append(obj_struct)\n\n return objects\n\n\ndef voc_ap(rec, prec, use_07_metric=False):\n \"\"\" ap = voc_ap(rec, prec, [use_07_metric])\n Compute VOC AP given precision and recall.\n If use_07_metric is true, uses the\n VOC 07 11 point method (default:False).\n \"\"\"\n if use_07_metric:\n # 11 point metric\n ap = 0.\n for t in np.arange(0., 1.1, 0.1):\n if np.sum(rec >= t) == 0:\n p = 0\n else:\n p = np.max(prec[rec >= t])\n ap = ap + p / 11.\n else:\n # correct AP calculation\n # first append sentinel values at the end\n mrec = np.concatenate(([0.], rec, [1.]))\n mpre = np.concatenate(([0.], prec, [0.]))\n\n # compute the precision envelope\n for i in range(mpre.size - 1, 0, -1):\n mpre[i - 1] = np.maximum(mpre[i - 1], mpre[i])\n\n # to calculate area under PR curve, look for points\n # where X axis (recall) changes value\n i = np.where(mrec[1:] != mrec[:-1])[0]\n\n # and sum (\\Delta recall) * prec\n ap = np.sum((mrec[i + 1] - mrec[i]) * mpre[i + 1])\n return ap\n\n\ndef voc_eval(detpath,\n annopath,\n imagesetfile,\n classname,\n cachedir,\n ovthresh=0.5,\n use_07_metric=False):\n \"\"\"rec, prec, ap = voc_eval(detpath,\n annopath,\n imagesetfile,\n classname,\n [ovthresh],\n [use_07_metric])\n\n Top level function that does the PASCAL VOC evaluation.\n\n detpath: Path to detections\n detpath.format(classname) should produce the detection results file.\n annopath: Path to annotations\n annopath.format(imagename) should be the xml annotations file.\n imagesetfile: Text file containing the list of images, one image per line.\n classname: Category name (duh)\n cachedir: Directory for caching the annotations\n [ovthresh]: Overlap threshold (default = 0.5)\n [use_07_metric]: Whether to use VOC07's 11 point AP computation\n (default False)\n \"\"\"\n # assumes detections are in detpath.format(classname)\n # assumes annotations are in annopath.format(imagename)\n # assumes imagesetfile is a text file with each line an image name\n # cachedir caches the annotations in a pickle file\n\n # first load gt\n if not os.path.isdir(cachedir):\n os.mkdir(cachedir)\n cachefile = os.path.join(cachedir, 'annots.pkl')\n\n # read list of images\n with open(imagesetfile, 'r') as f:\n lines = f.readlines()\n imagenames = [x.strip() for x in lines]\n\n if not os.path.isfile(cachefile):\n # load annotations\n recs = {}\n for i, imagename in enumerate(imagenames):\n recs[imagename] = parse_rec(annopath.format(imagename))\n if i % 100 == 0:\n print('Reading annotation for {:d}/{:d}'.format(\n i + 1, len(imagenames)))\n # save\n print('Saving cached annotations to {:s}'.format(cachefile))\n with open(cachefile, 'wb') as f:\n pickle.dump(recs, f)\n else:\n # load\n with open(cachefile, 'rb') as f:\n try:\n recs = pickle.load(f)\n except:\n recs = pickle.load(f, encoding='bytes')\n\n # extract gt objects for this class\n class_recs = {}\n npos = 0\n for imagename in imagenames:\n R = [obj for obj in recs[imagename] if obj['name'] == classname]\n bbox = np.array([x['bbox'] for x in R])\n difficult = np.array([x['difficult'] for x in R]).astype(np.bool)\n det = [False] * len(R)\n npos = npos + sum(~difficult)\n class_recs[imagename] = {'bbox': bbox,\n 'difficult': difficult,\n 'det': det}\n\n # read dets\n detfile = detpath.format(classname)\n with open(detfile, 'r') as f:\n lines = f.readlines()\n\n if len(lines) == 0:\n # No detection examples\n return 0, 0, 0, 0, npos\n\n splitlines = [x.strip().split(' ') for x in lines]\n image_ids = [x[0] for x in splitlines]\n confidence = np.array([float(x[1]) for x in splitlines])\n BB = np.array([[float(z) for z in x[2:]] for x in splitlines])\n\n nd = len(image_ids)\n tp = np.zeros(nd)\n fp = np.zeros(nd)\n\n if BB.shape[0] > 0:\n # sort by confidence\n sorted_ind = np.argsort(-confidence)\n sorted_scores = np.sort(-confidence)\n BB = BB[sorted_ind, :]\n image_ids = [image_ids[x] for x in sorted_ind]\n\n # go down dets and mark TPs and FPs\n for d in range(nd):\n R = class_recs[image_ids[d]]\n bb = BB[d, :].astype(float)\n ovmax = -np.inf\n BBGT = R['bbox'].astype(float)\n\n if BBGT.size > 0:\n # compute overlaps\n # intersection\n ixmin = np.maximum(BBGT[:, 0], bb[0])\n iymin = np.maximum(BBGT[:, 1], bb[1])\n ixmax = np.minimum(BBGT[:, 2], bb[2])\n iymax = np.minimum(BBGT[:, 3], bb[3])\n iw = np.maximum(ixmax - ixmin + 1., 0.)\n ih = np.maximum(iymax - iymin + 1., 0.)\n inters = iw * ih\n\n # union\n uni = ((bb[2] - bb[0] + 1.) * (bb[3] - bb[1] + 1.) +\n (BBGT[:, 2] - BBGT[:, 0] + 1.) *\n (BBGT[:, 3] - BBGT[:, 1] + 1.) - inters)\n\n overlaps = inters / uni\n ovmax = np.max(overlaps)\n jmax = np.argmax(overlaps)\n\n if ovmax > ovthresh:\n if not R['difficult'][jmax]:\n if not R['det'][jmax]:\n tp[d] = 1.\n R['det'][jmax] = 1\n else:\n fp[d] = 1.\n else:\n fp[d] = 1.\n\n # compute precision recall\n fp = np.cumsum(fp)\n tp = np.cumsum(tp)\n rec = tp / float(npos)\n # avoid divide by zero in case the first detection matches a difficult\n # ground truth\n prec = tp / np.maximum(tp + fp, np.finfo(np.float64).eps)\n ap = voc_ap(rec, prec, use_07_metric)\n\n return rec, prec, ap\n"
},
{
"path": "lib/model/__init__.py",
"content": ""
},
{
"path": "lib/model/faster_rcnn/__init__.py",
"content": ""
},
{
"path": "lib/model/faster_rcnn/faster_rcnn.py",
"content": "# --------------------------------------------------------\n# Pytorch Meta R-CNN\n# Written by Anny Xu, Xiaopeng Yan, based on the code from Jianwei Yang\n# --------------------------------------------------------\nimport random\nimport torch\nimport torch.nn as nn\nimport torch.nn.functional as F\nfrom torch.autograd import Variable\nimport torchvision.models as models\nfrom torch.autograd import Variable\nimport numpy as np\nfrom model.utils.config import cfg\nfrom model.rpn.rpn import _RPN\nfrom model.roi_pooling.modules.roi_pool import _RoIPooling\nfrom model.roi_crop.modules.roi_crop import _RoICrop\nfrom model.roi_align.modules.roi_align import RoIAlignAvg\nfrom model.rpn.proposal_target_layer_cascade import _ProposalTargetLayer\nimport time\nimport pdb\nfrom model.utils.net_utils import _smooth_l1_loss, _crop_pool_layer, _affine_grid_gen, _affine_theta\nimport pickle\n\n\nclass _fasterRCNN(nn.Module):\n \"\"\" faster RCNN \"\"\"\n\n def __init__(self, classes, class_agnostic, meta_train, meta_test=None, meta_loss=None):\n super(_fasterRCNN, self).__init__()\n self.classes = classes\n self.n_classes = len(classes)\n self.class_agnostic = class_agnostic\n self.meta_train = meta_train\n self.meta_test = meta_test\n self.meta_loss = meta_loss\n # loss\n self.RCNN_loss_cls = 0\n self.RCNN_loss_bbox = 0\n\n # define rpn\n self.RCNN_rpn = _RPN(self.dout_base_model)\n self.RCNN_proposal_target = _ProposalTargetLayer(self.n_classes)\n self.RCNN_roi_pool = _RoIPooling(cfg.POOLING_SIZE, cfg.POOLING_SIZE, 1.0 / 16.0)\n self.RCNN_roi_align = RoIAlignAvg(cfg.POOLING_SIZE, cfg.POOLING_SIZE, 1.0 / 16.0)\n\n self.grid_size = cfg.POOLING_SIZE * 2 if cfg.CROP_RESIZE_WITH_MAX_POOL else cfg.POOLING_SIZE\n self.RCNN_roi_crop = _RoICrop()\n\n\n def forward(self, im_data_list, im_info_list, gt_boxes_list, num_boxes_list, average_shot=None,\n mean_class_attentions=None):\n # return attentions for testing\n if average_shot:\n prn_data = im_data_list[0] # len(metaclass)*4*224*224\n attentions = self.prn_network(prn_data)\n return attentions\n # extract attentions for training\n if self.meta_train and self.training:\n prn_data = im_data_list[0] # len(metaclass)*4*224*224\n # feed prn data to prn_network\n attentions = self.prn_network(prn_data)\n prn_cls = im_info_list[0] # len(metaclass)\n\n im_data = im_data_list[-1]\n im_info = im_info_list[-1]\n gt_boxes = gt_boxes_list[-1]\n num_boxes = num_boxes_list[-1]\n\n batch_size = im_data.size(0)\n im_info = im_info.data\n gt_boxes = gt_boxes.data\n num_boxes = num_boxes.data\n\n # feed image data to base model to obtain base feature map\n base_feat = self.RCNN_base(self.rcnn_conv1(im_data))\n\n # feed base feature map tp RPN to obtain rois\n rois, rpn_loss_cls, rpn_loss_bbox = self.RCNN_rpn(base_feat, im_info, gt_boxes, num_boxes)\n\n # if it is training phase, then use ground truth bboxes for refining\n if self.training:\n roi_data = self.RCNN_proposal_target(rois, gt_boxes, num_boxes)\n rois, rois_label, rois_target, rois_inside_ws, rois_outside_ws = roi_data\n rois_label = Variable(rois_label.view(-1).long())\n rois_target = Variable(rois_target.view(-1, rois_target.size(2)))\n rois_inside_ws = Variable(rois_inside_ws.view(-1, rois_inside_ws.size(2)))\n rois_outside_ws = Variable(rois_outside_ws.view(-1, rois_outside_ws.size(2)))\n else:\n rois_label = None\n rois_target = None\n rois_inside_ws = None\n rois_outside_ws = None\n rpn_loss_cls = 0\n rpn_loss_bbox = 0\n\n rois = Variable(rois)\n\n # do roi pooling based on predicted rois\n if cfg.POOLING_MODE == 'crop':\n # pooled_feat_anchor = _crop_pool_layer(base_feat, rois.view(-1, 5))\n grid_xy = _affine_grid_gen(rois.view(-1, 5), base_feat.size()[2:], self.grid_size)\n grid_yx = torch.stack([grid_xy.data[:, :, :, 1], grid_xy.data[:, :, :, 0]], 3).contiguous()\n pooled_feat = self.RCNN_roi_crop(base_feat, Variable(grid_yx).detach())\n if cfg.CROP_RESIZE_WITH_MAX_POOL:\n pooled_feat = F.max_pool2d(pooled_feat, 2, 2)\n elif cfg.POOLING_MODE == 'align':\n pooled_feat = self.RCNN_roi_align(base_feat, rois.view(-1, 5)) # (b*128)*1024*7*7\n elif cfg.POOLING_MODE == 'pool':\n pooled_feat = self.RCNN_roi_pool(base_feat, rois.view(-1, 5))\n\n # feed pooled features to top model\n pooled_feat = self._head_to_tail(pooled_feat) # (b*128)*2048\n\n # meta training phase\n if self.meta_train:\n rcnn_loss_cls = []\n rcnn_loss_bbox = []\n # pooled feature maps need to operate channel-wise multiplication with the corresponding class's attentions of every roi of image\n for b in range(batch_size):\n zero = Variable(torch.FloatTensor([0]).cuda())\n proposal_labels = rois_label[b * 128:(b + 1) * 128].data.cpu().numpy()[0]\n unique_labels = list(np.unique(proposal_labels)) # the unique rois labels of the input image\n for i in range(attentions.size(0)): # attentions len(attentions)*2048\n if prn_cls[i].numpy()[0] + 1 not in unique_labels:\n rcnn_loss_cls.append(zero)\n rcnn_loss_bbox.append(zero)\n continue\n channel_wise_feat = pooled_feat[b * cfg.TRAIN.BATCH_SIZE:(b + 1) * cfg.TRAIN.BATCH_SIZE, :] * \\\n attentions[i] # 128x2048 channel-wise multiple\n bbox_pred = self.RCNN_bbox_pred(channel_wise_feat) # 128 * 4\n if self.training and not self.class_agnostic:\n # select the corresponding columns according to roi labels\n bbox_pred_view = bbox_pred.view(bbox_pred.size(0), int(bbox_pred.size(1) / 4), 4)\n bbox_pred_select = torch.gather(bbox_pred_view, 1,\n rois_label[\n b * cfg.TRAIN.BATCH_SIZE:(b + 1) * cfg.TRAIN.BATCH_SIZE].view(\n rois_label[b * cfg.TRAIN.BATCH_SIZE:(\n b + 1) * cfg.TRAIN.BATCH_SIZE].size(\n 0), 1, 1).expand(\n rois_label[b * cfg.TRAIN.BATCH_SIZE:(\n b + 1) * cfg.TRAIN.BATCH_SIZE].size(\n 0), 1,\n 4))\n bbox_pred = bbox_pred_select.squeeze(1)\n # compute object classification probability\n cls_score = self.RCNN_cls_score(channel_wise_feat) # 128 * 21\n\n if self.training:\n # classification loss\n RCNN_loss_cls = F.cross_entropy(cls_score, rois_label[b * 128:(b + 1) * 128])\n rcnn_loss_cls.append(RCNN_loss_cls)\n # bounding box regression L1 loss\n RCNN_loss_bbox = _smooth_l1_loss(bbox_pred, rois_target[b * 128:(b + 1) * 128],\n rois_inside_ws[b * 128:(b + 1) * 128],\n rois_outside_ws[b * 128:(b + 1) * 128])\n\n rcnn_loss_bbox.append(RCNN_loss_bbox)\n # meta attentions loss\n if self.meta_loss:\n attentions_score = self.Meta_cls_score(attentions)\n meta_loss = F.cross_entropy(attentions_score, Variable(torch.cat(prn_cls,dim=0).cuda()))\n else:\n meta_loss = 0\n\n return rois, rpn_loss_cls, rpn_loss_bbox, rcnn_loss_cls, rcnn_loss_bbox, rois_label, 0, 0, meta_loss\n\n elif self.meta_test:\n cls_prob_list = []\n bbox_pred_list = []\n for i in range(len(mean_class_attentions)):\n mean_attentions = mean_class_attentions[i]\n channel_wise_feat = pooled_feat * mean_attentions\n # compute bbox offset\n bbox_pred = self.RCNN_bbox_pred(channel_wise_feat)\n if self.training and not self.class_agnostic:\n # select the corresponding columns according to roi labels\n bbox_pred_view = bbox_pred.view(bbox_pred.size(0), int(bbox_pred.size(1) / 4), 4)\n bbox_pred_select = torch.gather(bbox_pred_view, 1,\n rois_label.view(rois_label.size(0), 1, 1).expand(rois_label.size(0),\n 1, 4))\n bbox_pred = bbox_pred_select.squeeze(1)\n\n # compute object classification probability\n cls_score = self.RCNN_cls_score(channel_wise_feat)\n cls_prob = F.softmax(cls_score)\n\n RCNN_loss_cls = 0\n RCNN_loss_bbox = 0\n\n if self.training:\n # classification loss\n RCNN_loss_cls = F.cross_entropy(cls_score, rois_label)\n # bounding box regression L1 loss\n RCNN_loss_bbox = _smooth_l1_loss(bbox_pred, rois_target, rois_inside_ws, rois_outside_ws)\n\n cls_prob = cls_prob.view(batch_size, rois.size(1), -1)\n bbox_pred = bbox_pred.view(batch_size, rois.size(1), -1)\n cls_prob_list.append(cls_prob)\n bbox_pred_list.append(bbox_pred)\n\n return rois, rpn_loss_cls, rpn_loss_bbox, RCNN_loss_cls, RCNN_loss_bbox, rois_label, cls_prob_list, bbox_pred_list, 0\n else:\n bbox_pred = self.RCNN_bbox_pred(pooled_feat)\n if self.training and not self.class_agnostic:\n # select the corresponding columns according to roi labels\n bbox_pred_view = bbox_pred.view(bbox_pred.size(0), int(bbox_pred.size(1) / 4), 4)\n bbox_pred_select = torch.gather(bbox_pred_view, 1,\n rois_label.view(rois_label.size(0), 1, 1).expand(rois_label.size(0), 1,\n 4))\n bbox_pred = bbox_pred_select.squeeze(1)\n\n # compute object classification probability\n cls_score = self.RCNN_cls_score(pooled_feat) # 128 * 1001\n cls_prob = F.softmax(cls_score)\n\n RCNN_loss_cls = 0\n RCNN_loss_bbox = 0\n\n if self.training:\n # classification loss\n RCNN_loss_cls = F.cross_entropy(cls_score, rois_label)\n\n # bounding box regression L1 loss\n RCNN_loss_bbox = _smooth_l1_loss(bbox_pred, rois_target, rois_inside_ws, rois_outside_ws)\n\n cls_prob = cls_prob.view(batch_size, rois.size(1), -1)\n bbox_pred = bbox_pred.view(batch_size, rois.size(1), -1)\n\n return rois, rpn_loss_cls, rpn_loss_bbox, RCNN_loss_cls, RCNN_loss_bbox, rois_label, cls_prob, bbox_pred, 0\n\n def _init_weights(self):\n def normal_init(m, mean, stddev, truncated=False):\n \"\"\"\n weight initalizer: truncated normal and random normal.\n \"\"\"\n # x is a parameter\n if truncated:\n m.weight.data.normal_().fmod_(2).mul_(stddev).add_(mean) # not a perfect approximation\n else:\n m.weight.data.normal_(mean, stddev)\n m.bias.data.zero_()\n\n normal_init(self.RCNN_rpn.RPN_Conv, 0, 0.01, cfg.TRAIN.TRUNCATED)\n normal_init(self.RCNN_rpn.RPN_cls_score, 0, 0.01, cfg.TRAIN.TRUNCATED)\n normal_init(self.RCNN_rpn.RPN_bbox_pred, 0, 0.01, cfg.TRAIN.TRUNCATED)\n normal_init(self.RCNN_cls_score, 0, 0.01, cfg.TRAIN.TRUNCATED)\n normal_init(self.RCNN_bbox_pred, 0, 0.001, cfg.TRAIN.TRUNCATED)\n\n def create_architecture(self):\n self._init_modules()\n self._init_weights()\n"
},
{
"path": "lib/model/faster_rcnn/resnet.py",
"content": "# --------------------------------------------------------\n# Pytorch Meta R-CNN\n# Written by Anny Xu, Xiaopeng Yan, based on code from Jianwei Yang\n# --------------------------------------------------------\nfrom __future__ import absolute_import\nfrom __future__ import division\nfrom __future__ import print_function\n\nfrom model.utils.config import cfg\nfrom model.faster_rcnn.faster_rcnn import _fasterRCNN\n\nimport torch\nimport torch.nn as nn\nimport torch.nn.functional as F\nfrom torch.autograd import Variable\nfrom torch.nn import init\nimport math\nimport torch.utils.model_zoo as model_zoo\nimport pdb\n\n__all__ = ['ResNet', 'resnet18', 'resnet34', 'resnet50', 'resnet101',\n 'resnet152']\n\n\nmodel_urls = {\n 'resnet18': 'https://s3.amazonaws.com/pytorch/models/resnet18-5c106cde.pth',\n 'resnet34': 'https://s3.amazonaws.com/pytorch/models/resnet34-333f7ec4.pth',\n 'resnet50': 'https://s3.amazonaws.com/pytorch/models/resnet50-19c8e357.pth',\n 'resnet101': 'https://s3.amazonaws.com/pytorch/models/resnet101-5d3b4d8f.pth',\n 'resnet152': 'https://s3.amazonaws.com/pytorch/models/resnet152-b121ed2d.pth',\n}\n\ndef conv3x3(in_planes, out_planes, stride=1):\n \"3x3 convolution with padding\"\n return nn.Conv2d(in_planes, out_planes, kernel_size=3, stride=stride,\n padding=1, bias=False)\n\ndef init_conv(conv,glu=True):\n init.xavier_uniform(conv.weight)\n if conv.bias is not None:\n conv.bias.data.zero_()\n\ndef init_linear(linear):\n init.constant(linear.weight,0)\n init.constant(linear.bias, 1)\n\nclass BasicBlock(nn.Module):\n expansion = 1\n\n def __init__(self, inplanes, planes, stride=1, downsample=None):\n super(BasicBlock, self).__init__()\n self.conv1 = conv3x3(inplanes, planes, stride)\n self.bn1 = nn.BatchNorm2d(planes)\n self.relu = nn.ReLU(inplace=True)\n self.conv2 = conv3x3(planes, planes)\n self.bn2 = nn.BatchNorm2d(planes)\n self.downsample = downsample\n self.stride = stride\n\n def forward(self, x):\n residual = x\n\n out = self.conv1(x)\n out = self.bn1(out)\n out = self.relu(out)\n\n out = self.conv2(out)\n out = self.bn2(out)\n\n if self.downsample is not None:\n residual = self.downsample(x)\n\n out += residual\n out = self.relu(out)\n\n return out\n\n\nclass Bottleneck(nn.Module):\n expansion = 4\n\n def __init__(self, inplanes, planes, stride=1, downsample=None):\n super(Bottleneck, self).__init__()\n self.conv1 = nn.Conv2d(inplanes, planes, kernel_size=1, stride=stride, bias=False) # change\n self.bn1 = nn.BatchNorm2d(planes)\n self.conv2 = nn.Conv2d(planes, planes, kernel_size=3, stride=1, # change\n padding=1, bias=False)\n self.bn2 = nn.BatchNorm2d(planes)\n self.conv3 = nn.Conv2d(planes, planes * 4, kernel_size=1, bias=False)\n self.bn3 = nn.BatchNorm2d(planes * 4)\n self.relu = nn.ReLU(inplace=True)\n self.downsample = downsample\n self.stride = stride\n\n def forward(self, x):\n residual = x\n\n out = self.conv1(x)\n out = self.bn1(out)\n out = self.relu(out)\n\n out = self.conv2(out)\n out = self.bn2(out)\n out = self.relu(out)\n\n out = self.conv3(out)\n out = self.bn3(out)\n\n if self.downsample is not None:\n residual = self.downsample(x)\n\n out += residual\n out = self.relu(out)\n\n return out\n\n\nclass ResNet(nn.Module):\n def __init__(self, block, layers, num_classes=1000):\n self.inplanes = 64\n super(ResNet, self).__init__()\n self.conv1 = nn.Conv2d(3, 64, kernel_size=7, stride=2, padding=3,\n bias=False)\n self.bn1 = nn.BatchNorm2d(64)\n self.relu = nn.ReLU(inplace=True)\n self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=0, ceil_mode=True) # change\n self.layer1 = self._make_layer(block, 64, layers[0])\n self.layer2 = self._make_layer(block, 128, layers[1], stride=2)\n self.layer3 = self._make_layer(block, 256, layers[2], stride=2)\n self.layer4 = self._make_layer(block, 512, layers[3], stride=2)\n # it is slightly better whereas slower to set stride = 1\n # self.layer4 = self._make_layer(block, 512, layers[3], stride=1)\n self.avgpool = nn.AvgPool2d(7)\n self.fc = nn.Linear(512 * block.expansion, num_classes)\n\n for m in self.modules():\n if isinstance(m, nn.Conv2d):\n n = m.kernel_size[0] * m.kernel_size[1] * m.out_channels\n m.weight.data.normal_(0, math.sqrt(2. / n))\n elif isinstance(m, nn.BatchNorm2d):\n m.weight.data.fill_(1)\n m.bias.data.zero_()\n\n def _make_layer(self, block, planes, blocks, stride=1):\n downsample = None\n if stride != 1 or self.inplanes != planes * block.expansion:\n downsample = nn.Sequential(\n nn.Conv2d(self.inplanes, planes * block.expansion,\n kernel_size=1, stride=stride, bias=False),\n nn.BatchNorm2d(planes * block.expansion),\n )\n\n layers = []\n layers.append(block(self.inplanes, planes, stride, downsample))\n self.inplanes = planes * block.expansion\n for i in range(1, blocks):\n layers.append(block(self.inplanes, planes))\n\n return nn.Sequential(*layers)\n\n def forward(self, x):\n x = self.conv1(x)\n x = self.bn1(x)\n x = self.relu(x)\n x = self.maxpool(x)\n\n x = self.layer1(x)\n x = self.layer2(x)\n x = self.layer3(x)\n x = self.layer4(x)\n\n x = self.avgpool(x)\n x = x.view(x.size(0), -1)\n x = self.fc(x)\n\n return x\n\n\ndef resnet18(pretrained=False):\n \"\"\"Constructs a ResNet-18 model.\n Args:\n pretrained (bool): If True, returns a model pre-trained on ImageNet\n \"\"\"\n model = ResNet(BasicBlock, [2, 2, 2, 2])\n if pretrained:\n model.load_state_dict(model_zoo.load_url(model_urls['resnet18']))\n return model\n\n\ndef resnet34(pretrained=False):\n \"\"\"Constructs a ResNet-34 model.\n Args:\n pretrained (bool): If True, returns a model pre-trained on ImageNet\n \"\"\"\n model = ResNet(BasicBlock, [3, 4, 6, 3])\n if pretrained:\n model.load_state_dict(model_zoo.load_url(model_urls['resnet34']))\n return model\n\n\ndef resnet50(pretrained=False):\n \"\"\"Constructs a ResNet-50 model.\n Args:\n pretrained (bool): If True, returns a model pre-trained on ImageNet\n \"\"\"\n model = ResNet(Bottleneck, [3, 4, 6, 3])\n if pretrained:\n model.load_state_dict(model_zoo.load_url(model_urls['resnet50']))\n return model\n\n\ndef resnet101(pretrained=False):\n \"\"\"Constructs a ResNet-101 model.\n Args:\n pretrained (bool): If True, returns a model pre-trained on ImageNet\n \"\"\"\n model = ResNet(Bottleneck, [3, 4, 23, 3])\n if pretrained:\n model.load_state_dict(model_zoo.load_url(model_urls['resnet101']))\n return model\n\n\ndef resnet152(pretrained=False):\n \"\"\"Constructs a ResNet-152 model.\n Args:\n pretrained (bool): If True, returns a model pre-trained on ImageNet\n \"\"\"\n model = ResNet(Bottleneck, [3, 8, 36, 3])\n if pretrained:\n model.load_state_dict(model_zoo.load_url(model_urls['resnet152']))\n return model\n\nclass resnet(_fasterRCNN):\n def __init__(self, classes, num_layers=101, pretrained=False, class_agnostic=False,meta_train=True,meta_test=None,meta_loss=None):\n self.model_path = 'data/pretrained_model/resnet101_caffe.pth'\n self.dout_base_model = 1024\n self.pretrained = pretrained\n self.class_agnostic = class_agnostic\n self.meta_train = meta_train\n self.meta_test = meta_test\n self.meta_loss = meta_loss\n\n _fasterRCNN.__init__(self, classes, class_agnostic,meta_train,meta_test,meta_loss)\n\n def _init_modules(self):\n resnet = resnet101()\n\n if self.pretrained == True:\n print(\"Loading pretrained weights from %s\" %(self.model_path))\n state_dict = torch.load(self.model_path)\n resnet.load_state_dict({k:v for k,v in state_dict.items() if k in resnet.state_dict()})\n\n # Build resnet.\n self.meta_conv1 = nn.Conv2d(4, 64, kernel_size=7, stride=2, padding=3, bias=False)\n self.rcnn_conv1 = resnet.conv1\n\n self.RCNN_base = nn.Sequential(resnet.bn1,resnet.relu,\n resnet.maxpool,resnet.layer1,resnet.layer2,resnet.layer3)\n\n self.RCNN_top = nn.Sequential(resnet.layer4)\n\n self.sigmoid = nn.Sigmoid()\n self.max_pooled = nn.MaxPool2d(2)\n\n self.RCNN_cls_score = nn.Linear(2048, self.n_classes)\n\n if self.meta_loss:\n self.Meta_cls_score = nn.Linear(2048, self.n_classes)\n\n if self.class_agnostic:\n self.RCNN_bbox_pred = nn.Linear(2048, 4) # x,y,w,h\n else:\n self.RCNN_bbox_pred = nn.Linear(2048, 4 * self.n_classes)\n\n\n # Fix blocks\n for p in self.rcnn_conv1.parameters(): p.requires_grad=False\n for p in self.RCNN_base[0].parameters(): p.requires_grad=False\n\n\n assert (0 <= cfg.RESNET.FIXED_BLOCKS < 5)\n if cfg.RESNET.FIXED_BLOCKS >= 4:\n for p in self.RCNN_top.parameters(): p.requires_grad = False\n if cfg.RESNET.FIXED_BLOCKS >= 3:\n for p in self.RCNN_base[5].parameters(): p.requires_grad=False\n if cfg.RESNET.FIXED_BLOCKS >= 2:\n for p in self.RCNN_base[4].parameters(): p.requires_grad=False\n if cfg.RESNET.FIXED_BLOCKS >= 1:\n for p in self.RCNN_base[3].parameters(): p.requires_grad=False\n\n def set_bn_fix(m):\n classname = m.__class__.__name__\n if classname.find('BatchNorm') != -1:\n for p in m.parameters(): p.requires_grad=False\n\n self.RCNN_base.apply(set_bn_fix)\n self.RCNN_top.apply(set_bn_fix)\n\n def train(self, mode=True):\n # Override train so that the training mode is set as we want\n nn.Module.train(self, mode)\n if mode:\n # Set fixed blocks to be in eval mode\n self.RCNN_base.eval()\n self.RCNN_base[4].train()\n self.RCNN_base[5].train()\n\n self.RCNN_base.eval()\n\n def set_bn_eval(m):\n classname = m.__class__.__name__\n if classname.find('BatchNorm') != -1:\n m.eval()\n\n self.RCNN_base.apply(set_bn_eval)\n self.RCNN_top.apply(set_bn_eval)\n\n def _head_to_tail(self, pool5):\n fc7 = self.RCNN_top(pool5).mean(3).mean(2)\n return fc7\n\n def prn_network(self,im_data):\n '''\n the Predictor-head Remodeling Network (PRN)\n :param im_data:\n :return attention vectors:\n '''\n base_feat = self.RCNN_base(self.meta_conv1(im_data))\n feature = self._head_to_tail(self.max_pooled(base_feat))\n attentions = self.sigmoid(feature)\n return attentions\n\n"
},
{
"path": "lib/model/faster_rcnn/trail.py",
"content": "# --------------------------------------------------------\n# Pytorch multi-GPU Faster R-CNN\n# Licensed under The MIT License [see LICENSE for details]\n# Written by Jiasen Lu, Jianwei Yang, based on code from Ross Girshick\n# --------------------------------------------------------\nimport _init_paths\nimport os\nimport sys\nimport numpy as np\nimport argparse\nimport pprint\nimport pdb\nimport time\n\nimport torch\nfrom torch.autograd import Variable\nimport torch.nn as nn\nimport torch.optim as optim\n\nimport torchvision.transforms as transforms\nfrom torch.utils.data.sampler import Sampler\n\nfrom roi_data_layer.roidb import combined_roidb\nfrom roi_data_layer.roibatchLoader import roibatchLoader\nfrom model.utils.config import cfg, cfg_from_file, cfg_from_list, get_output_dir\nfrom model.utils.net_utils import weights_normal_init, save_net, load_net, \\\n adjust_learning_rate, save_checkpoint, clip_gradient\n\nfrom model.faster_rcnn.vgg16 import vgg16\nfrom model.faster_rcnn.resnet_GNN import resnet\n\nimport pickle\n\ndef cout_w(prob):\n prob_weight = prob\n sum_value = np.sum(prob_weight)\n prob_weight = prob_weight / sum_value\n meam_value = np.mean(prob_weight)\n\n prob_weight = prob_weight - meam_value + 1\n prob_weight = np.log(prob_weight)\n return prob_weight\n\ndef cout_a(prob, NUM_ATTR_REL):\n prob_weight = prob[:, : NUM_ATTR_REL]\n sum_value = np.sum(prob_weight, keepdims=True, axis=1)\n zero_inds = np.nonzero(sum_value == 0)[0]\n sum_value[zero_inds,:] = 1.\n\n prob_weight = prob_weight / np.repeat(sum_value, prob_weight.shape[1], axis=1)\n\n meam_value = np.mean(prob_weight, keepdims=True, axis=1)\n meam_value = np.repeat(meam_value, prob_weight.shape[1], axis=1)\n prob_weight = prob_weight - meam_value + 1\n prob_weight = np.log(prob_weight)\n\n cls_cls_a = np.zeros((1001, 1001))\n for row in range(prob_weight.shape[0]):\n temp = np.zeros(1001)\n for col in range(prob_weight.shape[1]):\n xx = prob_weight[row, col] * prob_weight[:, col]\n temp += xx\n cls_cls_a[row] = temp\n pickle.dump(cls_cls_a, open('data/vg/clscls_attr.pkl', 'wb'))\n\n return cls_cls_a\n\ndef parse_args():\n \"\"\"\n Parse input arguments\n \"\"\"\n parser = argparse.ArgumentParser(description='Train a Fast R-CNN network')\n parser.add_argument('--dataset', dest='dataset',\n help='training dataset',\n default='vg', type=str)\n parser.add_argument('--net', dest='net',\n help='vgg16, res101',\n default='baseline', type=str)\n parser.add_argument('--start_epoch', dest='start_epoch',\n help='starting epoch',\n default=1, type=int)\n parser.add_argument('--epochs', dest='max_epochs',\n help='number of epochs to train',\n default=20, type=int)\n parser.add_argument('--disp_interval', dest='disp_interval',\n help='number of iterations to display',\n default=100, type=int)\n parser.add_argument('--checkpoint_interval', dest='checkpoint_interval',\n help='number of iterations to display',\n default=10000, type=int)\n\n parser.add_argument('--save_dir', dest='save_dir',\n help='directory to save models', default=\"exps/baseline/models\",\n nargs=argparse.REMAINDER)\n parser.add_argument('--nw', dest='num_workers',\n help='number of worker to load data',\n default=2, type=int)\n parser.add_argument('--cuda', dest='cuda',default=True, type=bool,\n help='whether use CUDA')\n parser.add_argument('--ls', dest='large_scale',\n help='whether use large imag scale',\n action='store_true')\n parser.add_argument('--ms', dest='multi_scale',\n help='whether to use multi scale training',\n action='store_true')\n parser.add_argument('--mGPUs', dest='mGPUs',\n help='whether use multiple GPUs',\n action='store_true')\n parser.add_argument('--bs', dest='batch_size',\n help='batch_size',\n default=2, type=int)\n parser.add_argument('--cag', dest='class_agnostic',default=False, type=bool,\n help='whether perform class_agnostic bbox regression')\n\n# config optimization\n parser.add_argument('--o', dest='optimizer',\n help='training optimizer',\n default=\"sgd\", type=str)\n parser.add_argument('--lr', dest='lr',\n help='starting learning rate',\n default=0.001, type=float)\n parser.add_argument('--lr_decay_step', dest='lr_decay_step',\n help='step to do learning rate decay, unit is epoch',\n default=4, type=int)\n parser.add_argument('--lr_decay_gamma', dest='lr_decay_gamma',\n help='learning rate decay ratio',\n default=0.1, type=float)\n\n# set training session\n parser.add_argument('--s', dest='session',\n help='training session',\n default=1, type=int)\n\n# resume trained model\n parser.add_argument('--r', dest='resume',\n help='resume checkpoint or not',\n default=False, type=bool)\n parser.add_argument('--checksession', dest='checksession',\n help='checksession to load model',\n default=1, type=int)\n parser.add_argument('--checkepoch', dest='checkepoch',\n help='checkepoch to load model',\n default=1, type=int)\n parser.add_argument('--checkpoint', dest='checkpoint',\n help='checkpoint to load model',\n default=0, type=int)\n# log and diaplay\n parser.add_argument('--use_tfboard', dest='use_tfboard',\n help='whether use tensorflow tensorboard',\n default=False, type=bool)\n parser.add_argument('--log_dir', dest='log_dir',\n help='directory to save logs', default='logs',\n type=str)\n\n args = parser.parse_args()\n return args\n\n\nclass sampler(Sampler):\n def __init__(self, train_size, batch_size):\n self.num_data = train_size\n self.num_per_batch = int(train_size / batch_size)\n self.batch_size = batch_size\n self.range = torch.arange(0,batch_size).view(1, batch_size).long()\n self.leftover_flag = False\n if train_size % batch_size:\n self.leftover = torch.arange(self.num_per_batch*batch_size, train_size).long()\n self.leftover_flag = True\n\n def __iter__(self):\n rand_num = torch.randperm(self.num_per_batch).view(-1,1) * self.batch_size\n self.rand_num = rand_num.expand(self.num_per_batch, self.batch_size) + self.range\n self.rand_num_view = self.rand_num.view(-1)\n\n if self.leftover_flag:\n self.rand_num_view = torch.cat((self.rand_num_view, self.leftover),0)\n\n return iter(self.rand_num_view)\n\n def __len__(self):\n return self.num_data\n\n\n args = parse_args()\n\n print('Called with args:')\n print(args)\n\n if args.use_tfboard:\n from model.utils.logger import Logger\n # Set the logger\n logger = Logger(args.log_dir)\n\n if args.dataset == \"pascal_voc\":\n args.imdb_name = \"voc_2007_trainval\"\n args.imdbval_name = \"voc_2007_test\"\n args.set_cfgs = ['ANCHOR_SCALES', '[8, 16, 32]', 'ANCHOR_RATIOS', '[0.5,1,2]', 'MAX_NUM_GT_BOXES', '20']\n elif args.dataset == \"pascal_voc_0712\":\n args.imdb_name = \"voc_2007_trainval+voc_2012_trainval\"\n args.imdbval_name = \"voc_2007_test\"\n args.set_cfgs = ['ANCHOR_SCALES', '[8, 16, 32]', 'ANCHOR_RATIOS', '[0.5,1,2]', 'MAX_NUM_GT_BOXES', '20']\n elif args.dataset == \"coco\":\n args.imdb_name = \"coco_2014_train+coco_2014_valminusminival\"\n args.imdbval_name = \"coco_2014_minival\"\n args.set_cfgs = ['ANCHOR_SCALES', '[4, 8, 16, 32]', 'ANCHOR_RATIOS', '[0.5,1,2]', 'MAX_NUM_GT_BOXES', '50']\n elif args.dataset == \"imagenet\":\n args.imdb_name = \"imagenet_train\"\n args.imdbval_name = \"imagenet_val\"\n args.set_cfgs = ['ANCHOR_SCALES', '[4, 8, 16, 32]', 'ANCHOR_RATIOS', '[0.5,1,2]', 'MAX_NUM_GT_BOXES', '30']\n elif args.dataset == \"vg\":\n # train sizes: train, smalltrain, minitrain\n # train scale: ['150-50-20', '150-50-50', '500-150-80', '750-250-150', '1750-700-450', '1600-400-20']\n args.imdb_name = \"vg_train\"\n args.imdbval_name = \"vg_val\"\n args.set_cfgs = ['ANCHOR_SCALES', '[2, 4, 8, 16, 32]', 'MAX_NUM_GT_BOXES', '50']\n # args.set_cfgs = ['ANCHOR_SCALES', '[4, 8, 16, 32]', 'ANCHOR_RATIOS', '[0.5,1,2]', 'MAX_NUM_GT_BOXES', '50']\n\n # args.cfg_file = \"cfgs/{}_ls.yml\".format(args.net) if args.large_scale else \"cfgs/{}.yml\".format(args.net)\n args.cfg_file = \"cfgs/res101_ms.yml\"#.format(args.net + \"_ms\" if args.multi_scale else \"\")\n\n if args.cfg_file is not None:\n cfg_from_file(args.cfg_file)\n if args.set_cfgs is not None:\n cfg_from_list(args.set_cfgs)\n\n print('Using config:')\n pprint.pprint(cfg)\n np.random.seed(cfg.RNG_SEED)\n\n #torch.backends.cudnn.benchmark = True\n if torch.cuda.is_available() and not args.cuda:\n print(\"WARNING: You have a CUDA device, so you should probably run with --cuda\")\n\n # train set\n # -- Note: Use validation set and disable the flipped to enable faster loading.\n cfg.TRAIN.USE_FLIPPED = True\n cfg.USE_GPU_NMS = args.cuda\n imdb, roidb, ratio_list, ratio_index = combined_roidb(args.imdb_name)\n train_size = len(roidb)\n\n print('{:d} roidb entries'.format(len(roidb)))\n sys.stdout.flush()\n\n output_dir = args.save_dir[0] + \"/\" + args.net + \"/\" + args.dataset\n if not os.path.exists(output_dir):\n os.makedirs(output_dir)\n\n # log_f = open(os.path.join(args.log_dir, time.strftime(\"%Y-%m-%d-%H:%M.txt\", time.localtime())), \"w\")\n # old_stdout = sys.stdout\n # sys.stdout = log_f\n\n sampler_batch = sampler(train_size, args.batch_size)\n\n dataset = roibatchLoader(roidb, ratio_list, ratio_index, args.batch_size, imdb.num_classes, training=True)\n\n dataloader = torch.utils.data.DataLoader(dataset, batch_size=args.batch_size,\n sampler=sampler_batch, num_workers=args.num_workers, pin_memory=False)\n\n # initilize the tensor holder here.\n im_data = torch.FloatTensor(1)\n im_info = torch.FloatTensor(1)\n num_boxes = torch.LongTensor(1)\n gt_boxes = torch.FloatTensor(1)\n# attr_prob = torch.FloatTensor(imdb._clscls_attr)\n # attr_prob = torch.FloatTensor(cout_a(imdb._class_to_attr, 200))\n# pair_prob = torch.FloatTensor(cout_w(imdb.pair))\n# pair_prob = pair_prob.unsqueeze(0).repeat(args.batch_size, 1, 1)\n\n\n # ship to cuda\n if args.cuda:\n im_data = im_data.cuda()\n im_info = im_info.cuda()\n num_boxes = num_boxes.cuda()\n gt_boxes = gt_boxes.cuda()\n\n\n # make variable\n im_data = Variable(im_data)\n im_info = Variable(im_info)\n num_boxes = Variable(num_boxes)\n gt_boxes = Variable(gt_boxes)\n# pair_prob = Variable(pair_prob)\n# attr_prob = Variable(attr_prob)\n\n if args.cuda:\n cfg.CUDA = True\n\n # initilize the network here.\n if args.net == 'baseline':\n fasterRCNN = resnet(imdb.classes, 101, pretrained=True, class_agnostic=args.class_agnostic)\n elif args.net == 'relationloss':\n fasterRCNN = resnet(imdb.classes, 101, pretrained=False, class_agnostic=args.class_agnostic, pair_prob=pair_prob)\n elif args.net == 'attributeloss':\n fasterRCNN = resnet(imdb.classes, 101, pretrained=True, class_agnostic=args.class_agnostic, attr_prob=attr_prob)\n\n\n fasterRCNN.create_architecture()\n\n lr = cfg.TRAIN.LEARNING_RATE\n lr = args.lr\n #tr_momentum = cfg.TRAIN.MOMENTUM\n #tr_momentum = args.momentum\n\n params = []\n for key, value in dict(fasterRCNN.named_parameters()).items():\n if value.requires_grad:\n if 'bias' in key:\n params += [{'params':[value],'lr':lr*(cfg.TRAIN.DOUBLE_BIAS + 1), \\\n 'weight_decay': cfg.TRAIN.BIAS_DECAY and cfg.TRAIN.WEIGHT_DECAY or 0}]\n else:\n params += [{'params':[value],'lr':lr, 'weight_decay': cfg.TRAIN.WEIGHT_DECAY}]\n\n if args.optimizer == \"adam\":\n lr = lr * 0.1\n optimizer = torch.optim.Adam(params)\n\n elif args.optimizer == \"sgd\":\n optimizer = torch.optim.SGD(params, momentum=cfg.TRAIN.MOMENTUM)\n\n if args.resume:\n load_name = os.path.join(output_dir,\n 'faster_rcnn_{}_{}_{}.pth'.format(args.checksession, args.checkepoch, args.checkpoint))\n print(\"loading checkpoint %s\" % (load_name))\n checkpoint = torch.load(load_name)\n args.session = checkpoint['session']\n args.start_epoch = checkpoint['epoch']\n fasterRCNN.load_state_dict(checkpoint['model'])\n optimizer.load_state_dict(checkpoint['optimizer'])\n lr = optimizer.param_groups[0]['lr']\n if 'pooling_mode' in checkpoint.keys():\n cfg.POOLING_MODE = checkpoint['pooling_mode']\n print(\"loaded checkpoint %s\" % (load_name))\n\n if args.mGPUs:\n fasterRCNN = nn.DataParallel(fasterRCNN)\n\n if args.cuda:\n fasterRCNN.cuda()\n\n iters_per_epoch = int(train_size / args.batch_size)\n\n\nfasterRCNN = resnet(imdb.classes, 101, pretrained=True, class_agnostic=args.class_agnostic)\nfasterRCNN.create_architecture()\nfasterRCNN.cuda()\nfasterRCNN.train()\ndata_iter = iter(dataloader)\n\ndata = next(data_iter)\nim_data.data.resize_(data[0].size()).copy_(data[0])\nim_info.data.resize_(data[1].size()).copy_(data[1])\ngt_boxes.data.resize_(data[2].size()).copy_(data[2])\nnum_boxes.data.resize_(data[3].size()).copy_(data[3])\n\nfasterRCNN.zero_grad()\n\n\n\nbatch_size = im_data.size(0)\n\nim_info = im_info.data\ngt_boxes = gt_boxes.data\nnum_boxes = num_boxes.data\n\n# feed image data to base model to obtain base feature map\nbase_feat = fasterRCNN.RCNN_base(im_data)\n\n# feed base feature map tp RPN to obtain rois\nrois, rpn_loss_cls, rpn_loss_bbox = fasterRCNN.RCNN_rpn(base_feat, im_info, gt_boxes, num_boxes)\n\n# if it is training phrase, then use ground trubut bboxes for refining\nif fasterRCNN.training:\n roi_data = fasterRCNN.RCNN_proposal_target(rois, gt_boxes, num_boxes)\n rois, rois_label, rois_target, rois_inside_ws, rois_outside_ws = roi_data\n\n rois_label = Variable(rois_label.view(-1).long())\n rois_target = Variable(rois_target.view(-1, rois_target.size(2)))\n rois_inside_ws = Variable(rois_inside_ws.view(-1, rois_inside_ws.size(2)))\n rois_outside_ws = Variable(rois_outside_ws.view(-1, rois_outside_ws.size(2)))\nelse:\n rois_label = None\n rois_target = None\n rois_inside_ws = None\n rois_outside_ws = None\n rpn_loss_cls = 0\n rpn_loss_bbox = 0\n\nrois = Variable(rois)\n# do roi pooling based on predicted rois\n\nif cfg.POOLING_MODE == 'crop':\n # pdb.set_trace()\n # pooled_feat_anchor = _crop_pool_layer(base_feat, rois.view(-1, 5))\n grid_xy = _affine_grid_gen(rois.view(-1, 5), base_feat.size()[2:], fasterRCNN.grid_size)\n grid_yx = torch.stack([grid_xy.data[:,:,:,1], grid_xy.data[:,:,:,0]], 3).contiguous()\n pooled_feat = fasterRCNN.RCNN_roi_crop(base_feat, Variable(grid_yx).detach())\n if cfg.CROP_RESIZE_WITH_MAX_POOL:\n pooled_feat = F.max_pool2d(pooled_feat, 2, 2)\nelif cfg.POOLING_MODE == 'align':\n pooled_feat = fasterRCNN.RCNN_roi_align(base_feat, rois.view(-1, 5))\nelif cfg.POOLING_MODE == 'pool':\n pooled_feat = fasterRCNN.RCNN_roi_pool(base_feat, rois.view(-1,5))"
},
{
"path": "lib/model/faster_rcnn/vgg16.py",
"content": "# --------------------------------------------------------\n# Tensorflow Faster R-CNN\n# Licensed under The MIT License [see LICENSE for details]\n# Written by Xinlei Chen\n# --------------------------------------------------------\nfrom __future__ import absolute_import\nfrom __future__ import division\nfrom __future__ import print_function\n\nimport torch\nimport torch.nn as nn\nimport torch.nn.functional as F\nfrom torch.autograd import Variable\nimport math\nimport torchvision.models as models\nfrom model.faster_rcnn.faster_rcnn import _fasterRCNN\nimport pdb\n\nclass vgg16(_fasterRCNN):\n def __init__(self, classes, pretrained=False, class_agnostic=False):\n self.model_path = 'data/pretrained_model/vgg16_caffe.pth'\n self.dout_base_model = 512\n self.pretrained = pretrained\n self.class_agnostic = class_agnostic\n\n _fasterRCNN.__init__(self, classes, class_agnostic)\n\n def _init_modules(self):\n vgg = models.vgg16()\n if self.pretrained:\n print(\"Loading pretrained weights from %s\" %(self.model_path))\n state_dict = torch.load(self.model_path)\n vgg.load_state_dict({k:v for k,v in state_dict.items() if k in vgg.state_dict()})\n\n vgg.classifier = nn.Sequential(*list(vgg.classifier._modules.values())[:-1])\n\n # not using the last maxpool layer\n self.RCNN_base = nn.Sequential(*list(vgg.features._modules.values())[:-1])\n\n # Fix the layers before conv3:\n for layer in range(10):\n for p in self.RCNN_base[layer].parameters(): p.requires_grad = False\n\n # self.RCNN_base = _RCNN_base(vgg.features, self.classes, self.dout_base_model)\n\n self.RCNN_top = vgg.classifier\n\n # not using the last maxpool layer\n self.RCNN_cls_score = nn.Linear(4096, self.n_classes)\n\n if self.class_agnostic:\n self.RCNN_bbox_pred = nn.Linear(4096, 4)\n else:\n self.RCNN_bbox_pred = nn.Linear(4096, 4 * self.n_classes) \n\n def _head_to_tail(self, pool5):\n \n pool5_flat = pool5.view(pool5.size(0), -1)\n fc7 = self.RCNN_top(pool5_flat)\n\n return fc7\n\n"
},
{
"path": "lib/model/nms/.gitignore",
"content": "*.c\n*.cpp\n*.so\n"
},
{
"path": "lib/model/nms/__init__.py",
"content": ""
},
{
"path": "lib/model/nms/_ext/__init__.py",
"content": ""
},
{
"path": "lib/model/nms/_ext/nms/__init__.py",
"content": "\nfrom torch.utils.ffi import _wrap_function\nfrom ._nms import lib as _lib, ffi as _ffi\n\n__all__ = []\ndef _import_symbols(locals):\n for symbol in dir(_lib):\n fn = getattr(_lib, symbol)\n if callable(fn):\n locals[symbol] = _wrap_function(fn, _ffi)\n else:\n locals[symbol] = fn\n __all__.append(symbol)\n\n_import_symbols(locals())\n"
},
{
"path": "lib/model/nms/build.py",
"content": "from __future__ import print_function\nimport os\nimport torch\nfrom torch.utils.ffi import create_extension\n\n#this_file = os.path.dirname(__file__)\n\nsources = []\nheaders = []\ndefines = []\nwith_cuda = False\n\nif torch.cuda.is_available():\n print('Including CUDA code.')\n sources += ['src/nms_cuda.c']\n headers += ['src/nms_cuda.h']\n defines += [('WITH_CUDA', None)]\n with_cuda = True\n\nthis_file = os.path.dirname(os.path.realpath(__file__))\nprint(this_file)\nextra_objects = ['src/nms_cuda_kernel.cu.o']\nextra_objects = [os.path.join(this_file, fname) for fname in extra_objects]\nprint(extra_objects)\n\nffi = create_extension(\n '_ext.nms',\n headers=headers,\n sources=sources,\n define_macros=defines,\n relative_to=__file__,\n with_cuda=with_cuda,\n extra_objects=extra_objects\n)\n\nif __name__ == '__main__':\n ffi.build()\n"
},
{
"path": "lib/model/nms/nms_cpu.py",
"content": "from __future__ import absolute_import\n\nimport numpy as np\nimport torch\n\ndef nms_cpu(dets, thresh):\n dets = dets.cpu().numpy()\n x1 = dets[:, 0]\n y1 = dets[:, 1]\n x2 = dets[:, 2]\n y2 = dets[:, 3]\n scores = dets[:, 4]\n\n areas = (x2 - x1 + 1) * (y2 - y1 + 1)\n order = scores.argsort()[::-1]\n\n keep = []\n while order.size > 0:\n i = order.item(0)\n keep.append(i)\n xx1 = np.maximum(x1[i], x1[order[1:]])\n yy1 = np.maximum(y1[i], y1[order[1:]])\n xx2 = np.minimum(x2[i], x2[order[1:]])\n yy2 = np.minimum(y2[i], y2[order[1:]])\n\n w = np.maximum(0.0, xx2 - xx1 + 1)\n h = np.maximum(0.0, yy2 - yy1 + 1)\n inter = w * h\n ovr = inter / (areas[i] + areas[order[1:]] - inter)\n\n inds = np.where(ovr <= thresh)[0]\n order = order[inds + 1]\n\n return torch.IntTensor(keep)\n\n\n\n\ndef nms_cpu_np(dets, thresh):\n x1 = dets[:, 0]\n y1 = dets[:, 1]\n x2 = dets[:, 2]\n y2 = dets[:, 3]\n scores = dets[:, 4]\n\n areas = (x2 - x1 + 1) * (y2 - y1 + 1)\n order = scores.argsort()[::-1]\n\n keep = []\n while order.size > 0:\n i = order.item(0)\n keep.append(i)\n xx1 = np.maximum(x1[i], x1[order[1:]])\n yy1 = np.maximum(y1[i], y1[order[1:]])\n xx2 = np.minimum(x2[i], x2[order[1:]])\n yy2 = np.minimum(y2[i], y2[order[1:]])\n\n w = np.maximum(0.0, xx2 - xx1 + 1)\n h = np.maximum(0.0, yy2 - yy1 + 1)\n inter = w * h\n ovr = inter / (areas[i] + areas[order[1:]] - inter)\n\n inds = np.where(ovr <= thresh)[0]\n order = order[inds + 1]\n\n return keep\n\n\n\ndef soft_nms_cpu(dets, threshold=0.001, Nt=0.3, method=1):\n boxes = dets.cpu().numpy()\n N = dets.shape[0]\n pos = 0\n maxscore = 0\n maxpos = 0\n \n for i in range(N):\n maxscore = boxes[i, 4]\n maxpos = i\n \n tx1 = boxes[i,0]\n ty1 = boxes[i,1]\n tx2 = boxes[i,2]\n ty2 = boxes[i,3]\n ts = boxes[i,4]\n \n pos = i + 1\n # get max box\n while pos < N:\n if maxscore < boxes[pos, 4]:\n maxscore = boxes[pos, 4]\n maxpos = pos\n pos = pos + 1\n \n # add max box as a detection\n boxes[i,0] = boxes[maxpos,0]\n boxes[i,1] = boxes[maxpos,1]\n boxes[i,2] = boxes[maxpos,2]\n boxes[i,3] = boxes[maxpos,3]\n boxes[i,4] = boxes[maxpos,4]\n \n # swap ith box with position of max box\n boxes[maxpos,0] = tx1\n boxes[maxpos,1] = ty1\n boxes[maxpos,2] = tx2\n boxes[maxpos,3] = ty2\n boxes[maxpos,4] = ts\n \n tx1 = boxes[i,0]\n ty1 = boxes[i,1]\n tx2 = boxes[i,2]\n ty2 = boxes[i,3]\n ts = boxes[i,4]\n \n pos = i + 1\n # NMS iterations, note that N changes if detection boxes fall below threshold\n while pos < N:\n x1 = boxes[pos, 0]\n y1 = boxes[pos, 1]\n x2 = boxes[pos, 2]\n y2 = boxes[pos, 3]\n s = boxes[pos, 4]\n \n area = (x2 - x1 + 1) * (y2 - y1 + 1)\n iw = (min(tx2, x2) - max(tx1, x1) + 1)\n if iw > 0:\n ih = (min(ty2, y2) - max(ty1, y1) + 1)\n if ih > 0:\n ua = float((tx2 - tx1 + 1) * (ty2 - ty1 + 1) + area - iw * ih)\n ov = iw * ih / ua #iou between max box and detection box\n \n if method == 1: # linear\n if ov > Nt: \n weight = 1 - ov\n else:\n weight = 1\n elif method == 2: # gaussian\n weight = np.exp(-(ov * ov)/sigma)\n else: # original NMS\n if ov > Nt: \n weight = 0\n else:\n weight = 1\n \n boxes[pos, 4] = weight*boxes[pos, 4]\n \t\t \n \t\t # if box score falls below threshold, discard the box by swapping with last box\n \t\t # update N\n if boxes[pos, 4] < threshold:\n boxes[pos,0] = boxes[N-1, 0]\n boxes[pos,1] = boxes[N-1, 1]\n boxes[pos,2] = boxes[N-1, 2]\n boxes[pos,3] = boxes[N-1, 3]\n boxes[pos,4] = boxes[N-1, 4]\n N = N - 1\n pos = pos - 1\n \n pos = pos + 1\n\n keep = [i for i in range(N)]\n return keep, boxes\n \n\ndef nms_domain(dets, dets_small, thresh_small=0.85, thresh_big=0.5):\n# dets = dets.cpu().numpy()\n# dets_small = dets_small.cpu().numpy()\n x1 = dets[:, 0]\n y1 = dets[:, 1]\n x2 = dets[:, 2]\n y2 = dets[:, 3]\n scores = dets[:, 4]\n\n x21 = dets_small[:, 0]\n y21 = dets_small[:, 1]\n x22 = dets_small[:, 2]\n y22 = dets_small[:, 3]\n scores2 = dets_small[:, 4]\n\n areas = (x2 - x1 + 1) * (y2 - y1 + 1)\n order = scores.argsort()[::-1]\n\n areas2 = (x22 - x21 + 1) * (y22 - y21 + 1)\n order2 = scores2.argsort()[::-1]\n \n throw = set()\n keep = set(list(range(len(dets_small))))\n for i in range(len(dets)):\n xx1 = np.maximum(x1[i], x21)\n yy1 = np.maximum(y1[i], y21)\n xx2 = np.minimum(x2[i], x22)\n yy2 = np.minimum(y2[i], y22)\n\n w = np.maximum(0.0, xx2 - xx1 + 1)\n h = np.maximum(0.0, yy2 - yy1 + 1)\n inter = w * h\n\n ovr_1 = inter / (areas[i])\n ovr_2 = inter / (areas2)\n\n throw_array = np.where((ovr_2 > thresh_small) & (ovr_1 < thresh_big))[0].tolist()\n throw.update(throw_array)\n keep = list(keep - throw)\n return keep\n \n "
},
{
"path": "lib/model/nms/nms_gpu.py",
"content": "from __future__ import absolute_import\nimport torch\nimport numpy as np\nfrom ._ext import nms\nimport pdb\n\ndef nms_gpu(dets, thresh):\n\tkeep = dets.new(dets.size(0), 1).zero_().int()\n\tnum_out = dets.new(1).zero_().int()\n\tnms.nms_cuda(keep, dets, num_out, thresh)\n\tkeep = keep[:num_out[0]]\n\treturn keep\n"
},
{
"path": "lib/model/nms/nms_kernel.cu",
"content": "// ------------------------------------------------------------------\n// Faster R-CNN\n// Copyright (c) 2015 Microsoft\n// Licensed under The MIT License [see fast-rcnn/LICENSE for details]\n// Written by Shaoqing Ren\n// ------------------------------------------------------------------\n\n#include \"gpu_nms.hpp\"\n#include \n#include \n\n#define CUDA_CHECK(condition) \\\n /* Code block avoids redefinition of cudaError_t error */ \\\n do { \\\n cudaError_t error = condition; \\\n if (error != cudaSuccess) { \\\n std::cout << cudaGetErrorString(error) << std::endl; \\\n } \\\n } while (0)\n\n#define DIVUP(m,n) ((m) / (n) + ((m) % (n) > 0))\nint const threadsPerBlock = sizeof(unsigned long long) * 8;\n\n__device__ inline float devIoU(float const * const a, float const * const b) {\n float left = max(a[0], b[0]), right = min(a[2], b[2]);\n float top = max(a[1], b[1]), bottom = min(a[3], b[3]);\n float width = max(right - left + 1, 0.f), height = max(bottom - top + 1, 0.f);\n float interS = width * height;\n float Sa = (a[2] - a[0] + 1) * (a[3] - a[1] + 1);\n float Sb = (b[2] - b[0] + 1) * (b[3] - b[1] + 1);\n return interS / (Sa + Sb - interS);\n}\n\n__global__ void nms_kernel(const int n_boxes, const float nms_overlap_thresh,\n const float *dev_boxes, unsigned long long *dev_mask) {\n const int row_start = blockIdx.y;\n const int col_start = blockIdx.x;\n\n // if (row_start > col_start) return;\n\n const int row_size =\n min(n_boxes - row_start * threadsPerBlock, threadsPerBlock);\n const int col_size =\n min(n_boxes - col_start * threadsPerBlock, threadsPerBlock);\n\n __shared__ float block_boxes[threadsPerBlock * 5];\n if (threadIdx.x < col_size) {\n block_boxes[threadIdx.x * 5 + 0] =\n dev_boxes[(threadsPerBlock * col_start + threadIdx.x) * 5 + 0];\n block_boxes[threadIdx.x * 5 + 1] =\n dev_boxes[(threadsPerBlock * col_start + threadIdx.x) * 5 + 1];\n block_boxes[threadIdx.x * 5 + 2] =\n dev_boxes[(threadsPerBlock * col_start + threadIdx.x) * 5 + 2];\n block_boxes[threadIdx.x * 5 + 3] =\n dev_boxes[(threadsPerBlock * col_start + threadIdx.x) * 5 + 3];\n block_boxes[threadIdx.x * 5 + 4] =\n dev_boxes[(threadsPerBlock * col_start + threadIdx.x) * 5 + 4];\n }\n __syncthreads();\n\n if (threadIdx.x < row_size) {\n const int cur_box_idx = threadsPerBlock * row_start + threadIdx.x;\n const float *cur_box = dev_boxes + cur_box_idx * 5;\n int i = 0;\n unsigned long long t = 0;\n int start = 0;\n if (row_start == col_start) {\n start = threadIdx.x + 1;\n }\n for (i = start; i < col_size; i++) {\n if (devIoU(cur_box, block_boxes + i * 5) > nms_overlap_thresh) {\n t |= 1ULL << i;\n }\n }\n const int col_blocks = DIVUP(n_boxes, threadsPerBlock);\n dev_mask[cur_box_idx * col_blocks + col_start] = t;\n }\n}\n\nvoid _set_device(int device_id) {\n int current_device;\n CUDA_CHECK(cudaGetDevice(¤t_device));\n if (current_device == device_id) {\n return;\n }\n // The call to cudaSetDevice must come before any calls to Get, which\n // may perform initialization using the GPU.\n CUDA_CHECK(cudaSetDevice(device_id));\n}\n\nvoid _nms(int* keep_out, int* num_out, const float* boxes_host, int boxes_num,\n int boxes_dim, float nms_overlap_thresh, int device_id) {\n _set_device(device_id);\n\n float* boxes_dev = NULL;\n unsigned long long* mask_dev = NULL;\n\n const int col_blocks = DIVUP(boxes_num, threadsPerBlock);\n\n CUDA_CHECK(cudaMalloc(&boxes_dev,\n boxes_num * boxes_dim * sizeof(float)));\n CUDA_CHECK(cudaMemcpy(boxes_dev,\n boxes_host,\n boxes_num * boxes_dim * sizeof(float),\n cudaMemcpyHostToDevice));\n\n CUDA_CHECK(cudaMalloc(&mask_dev,\n boxes_num * col_blocks * sizeof(unsigned long long)));\n\n dim3 blocks(DIVUP(boxes_num, threadsPerBlock),\n DIVUP(boxes_num, threadsPerBlock));\n dim3 threads(threadsPerBlock);\n nms_kernel<<>>(boxes_num,\n nms_overlap_thresh,\n boxes_dev,\n mask_dev);\n\n std::vector mask_host(boxes_num * col_blocks);\n CUDA_CHECK(cudaMemcpy(&mask_host[0],\n mask_dev,\n sizeof(unsigned long long) * boxes_num * col_blocks,\n cudaMemcpyDeviceToHost));\n\n std::vector remv(col_blocks);\n memset(&remv[0], 0, sizeof(unsigned long long) * col_blocks);\n\n int num_to_keep = 0;\n for (int i = 0; i < boxes_num; i++) {\n int nblock = i / threadsPerBlock;\n int inblock = i % threadsPerBlock;\n\n if (!(remv[nblock] & (1ULL << inblock))) {\n keep_out[num_to_keep++] = i;\n unsigned long long *p = &mask_host[0] + i * col_blocks;\n for (int j = nblock; j < col_blocks; j++) {\n remv[j] |= p[j];\n }\n }\n }\n *num_out = num_to_keep;\n\n CUDA_CHECK(cudaFree(boxes_dev));\n CUDA_CHECK(cudaFree(mask_dev));\n}\n"
},
{
"path": "lib/model/nms/nms_wrapper.py",
"content": "# --------------------------------------------------------\n# Fast R-CNN\n# Copyright (c) 2015 Microsoft\n# Licensed under The MIT License [see LICENSE for details]\n# Written by Ross Girshick\n# --------------------------------------------------------\nimport torch\nfrom model.utils.config import cfg\nif torch.cuda.is_available():\n from model.nms.nms_gpu import nms_gpu\nfrom model.nms.nms_cpu import nms_cpu\n\ndef nms(dets, thresh, force_cpu=False):\n \"\"\"Dispatch to either CPU or GPU NMS implementations.\"\"\"\n if dets.shape[0] == 0:\n return []\n # ---numpy version---\n # original: return gpu_nms(dets, thresh, device_id=cfg.GPU_ID)\n # ---pytorch version---\n\n return nms_gpu(dets, thresh) if force_cpu == False else nms_cpu(dets, thresh)\n"
},
{
"path": "lib/model/nms/src/nms_cuda.h",
"content": "// int nms_cuda(THCudaTensor *keep_out, THCudaTensor *num_out,\n// THCudaTensor *boxes_host, THCudaTensor *nms_overlap_thresh);\n\nint nms_cuda(THCudaIntTensor *keep_out, THCudaTensor *boxes_host,\n THCudaIntTensor *num_out, float nms_overlap_thresh);\n"
},
{
"path": "lib/model/nms/src/nms_cuda_kernel.cu",
"content": "// ------------------------------------------------------------------\n// Faster R-CNN\n// Copyright (c) 2015 Microsoft\n// Licensed under The MIT License [see fast-rcnn/LICENSE for details]\n// Written by Shaoqing Ren\n// ------------------------------------------------------------------\n\n#include \n#include \n#include \n#include \n#include \"nms_cuda_kernel.h\"\n\n#define CUDA_WARN(XXX) \\\n do { if (XXX != cudaSuccess) std::cout << \"CUDA Error: \" << \\\n cudaGetErrorString(XXX) << \", at line \" << __LINE__ \\\n<< std::endl; cudaDeviceSynchronize(); } while (0)\n\n#define CUDA_CHECK(condition) \\\n /* Code block avoids redefinition of cudaError_t error */ \\\n do { \\\n cudaError_t error = condition; \\\n if (error != cudaSuccess) { \\\n std::cout << cudaGetErrorString(error) << std::endl; \\\n } \\\n } while (0)\n\n#define DIVUP(m,n) ((m) / (n) + ((m) % (n) > 0))\nint const threadsPerBlock = sizeof(unsigned long long) * 8;\n\n__device__ inline float devIoU(float const * const a, float const * const b) {\n float left = max(a[0], b[0]), right = min(a[2], b[2]);\n float top = max(a[1], b[1]), bottom = min(a[3], b[3]);\n float width = max(right - left + 1, 0.f), height = max(bottom - top + 1, 0.f);\n float interS = width * height;\n float Sa = (a[2] - a[0] + 1) * (a[3] - a[1] + 1);\n float Sb = (b[2] - b[0] + 1) * (b[3] - b[1] + 1);\n return interS / (Sa + Sb - interS);\n}\n\n__global__ void nms_kernel(int n_boxes, float nms_overlap_thresh,\n float *dev_boxes, unsigned long long *dev_mask) {\n const int row_start = blockIdx.y;\n const int col_start = blockIdx.x;\n\n // if (row_start > col_start) return;\n\n const int row_size =\n min(n_boxes - row_start * threadsPerBlock, threadsPerBlock);\n const int col_size =\n min(n_boxes - col_start * threadsPerBlock, threadsPerBlock);\n\n __shared__ float block_boxes[threadsPerBlock * 5];\n if (threadIdx.x < col_size) {\n block_boxes[threadIdx.x * 5 + 0] =\n dev_boxes[(threadsPerBlock * col_start + threadIdx.x) * 5 + 0];\n block_boxes[threadIdx.x * 5 + 1] =\n dev_boxes[(threadsPerBlock * col_start + threadIdx.x) * 5 + 1];\n block_boxes[threadIdx.x * 5 + 2] =\n dev_boxes[(threadsPerBlock * col_start + threadIdx.x) * 5 + 2];\n block_boxes[threadIdx.x * 5 + 3] =\n dev_boxes[(threadsPerBlock * col_start + threadIdx.x) * 5 + 3];\n block_boxes[threadIdx.x * 5 + 4] =\n dev_boxes[(threadsPerBlock * col_start + threadIdx.x) * 5 + 4];\n }\n __syncthreads();\n\n if (threadIdx.x < row_size) {\n const int cur_box_idx = threadsPerBlock * row_start + threadIdx.x;\n const float *cur_box = dev_boxes + cur_box_idx * 5;\n int i = 0;\n unsigned long long t = 0;\n int start = 0;\n if (row_start == col_start) {\n start = threadIdx.x + 1;\n }\n for (i = start; i < col_size; i++) {\n if (devIoU(cur_box, block_boxes + i * 5) > nms_overlap_thresh) {\n t |= 1ULL << i;\n }\n }\n const int col_blocks = DIVUP(n_boxes, threadsPerBlock);\n dev_mask[cur_box_idx * col_blocks + col_start] = t;\n }\n}\n\nvoid nms_cuda_compute(int* keep_out, int *num_out, float* boxes_host, int boxes_num,\n int boxes_dim, float nms_overlap_thresh) {\n\n float* boxes_dev = NULL;\n unsigned long long* mask_dev = NULL;\n\n const int col_blocks = DIVUP(boxes_num, threadsPerBlock);\n\n CUDA_CHECK(cudaMalloc(&boxes_dev,\n boxes_num * boxes_dim * sizeof(float)));\n CUDA_CHECK(cudaMemcpy(boxes_dev,\n boxes_host,\n boxes_num * boxes_dim * sizeof(float),\n cudaMemcpyHostToDevice));\n\n CUDA_CHECK(cudaMalloc(&mask_dev,\n boxes_num * col_blocks * sizeof(unsigned long long)));\n\n dim3 blocks(DIVUP(boxes_num, threadsPerBlock),\n DIVUP(boxes_num, threadsPerBlock));\n dim3 threads(threadsPerBlock);\n\n // printf(\"i am at line %d\\n\", boxes_num);\n // printf(\"i am at line %d\\n\", boxes_dim); \n\n nms_kernel<<>>(boxes_num,\n nms_overlap_thresh,\n boxes_dev,\n mask_dev);\n\n std::vector mask_host(boxes_num * col_blocks);\n CUDA_CHECK(cudaMemcpy(&mask_host[0],\n mask_dev,\n sizeof(unsigned long long) * boxes_num * col_blocks,\n cudaMemcpyDeviceToHost));\n\n std::vector remv(col_blocks);\n memset(&remv[0], 0, sizeof(unsigned long long) * col_blocks);\n\n // we need to create a memory for keep_out on cpu\n // otherwise, the following code cannot run\n\n int* keep_out_cpu = new int[boxes_num];\n\n int num_to_keep = 0;\n for (int i = 0; i < boxes_num; i++) {\n int nblock = i / threadsPerBlock;\n int inblock = i % threadsPerBlock;\n\n if (!(remv[nblock] & (1ULL << inblock))) {\n // orignal: keep_out[num_to_keep++] = i;\n keep_out_cpu[num_to_keep++] = i;\n unsigned long long *p = &mask_host[0] + i * col_blocks;\n for (int j = nblock; j < col_blocks; j++) {\n remv[j] |= p[j];\n }\n }\n }\n\n // copy keep_out_cpu to keep_out on gpu\n CUDA_WARN(cudaMemcpy(keep_out, keep_out_cpu, boxes_num * sizeof(int),cudaMemcpyHostToDevice)); \n\n // *num_out = num_to_keep;\n\n // original: *num_out = num_to_keep;\n // copy num_to_keep to num_out on gpu\n\n CUDA_WARN(cudaMemcpy(num_out, &num_to_keep, 1 * sizeof(int),cudaMemcpyHostToDevice)); \n\n // release cuda memory\n CUDA_CHECK(cudaFree(boxes_dev));\n CUDA_CHECK(cudaFree(mask_dev));\n // release cpu memory\n delete []keep_out_cpu;\n}\n"
},
{
"path": "lib/model/nms/src/nms_cuda_kernel.h",
"content": "#ifdef __cplusplus\nextern \"C\" {\n#endif\n\nvoid nms_cuda_compute(int* keep_out, int *num_out, float* boxes_host, int boxes_num,\n int boxes_dim, float nms_overlap_thresh);\n\n#ifdef __cplusplus\n}\n#endif\n"
},
{
"path": "lib/model/roi_align/__init__.py",
"content": ""
},
{
"path": "lib/model/roi_align/_ext/__init__.py",
"content": ""
},
{
"path": "lib/model/roi_align/_ext/roi_align/__init__.py",
"content": "\nfrom torch.utils.ffi import _wrap_function\nfrom ._roi_align import lib as _lib, ffi as _ffi\n\n__all__ = []\ndef _import_symbols(locals):\n for symbol in dir(_lib):\n fn = getattr(_lib, symbol)\n if callable(fn):\n locals[symbol] = _wrap_function(fn, _ffi)\n else:\n locals[symbol] = fn\n __all__.append(symbol)\n\n_import_symbols(locals())\n"
},
{
"path": "lib/model/roi_align/build.py",
"content": "from __future__ import print_function\nimport os\nimport torch\nfrom torch.utils.ffi import create_extension\n\n# sources = ['src/roi_align.c']\n# headers = ['src/roi_align.h']\nsources = []\nheaders = []\ndefines = []\nwith_cuda = False\n\nif torch.cuda.is_available():\n print('Including CUDA code.')\n sources += ['src/roi_align_cuda.c']\n headers += ['src/roi_align_cuda.h']\n defines += [('WITH_CUDA', None)]\n with_cuda = True\n\nthis_file = os.path.dirname(os.path.realpath(__file__))\nprint(this_file)\nextra_objects = ['src/roi_align_kernel.cu.o']\nextra_objects = [os.path.join(this_file, fname) for fname in extra_objects]\n\nffi = create_extension(\n '_ext.roi_align',\n headers=headers,\n sources=sources,\n define_macros=defines,\n relative_to=__file__,\n with_cuda=with_cuda,\n extra_objects=extra_objects\n)\n\nif __name__ == '__main__':\n ffi.build()\n"
},
{
"path": "lib/model/roi_align/functions/__init__.py",
"content": ""
},
{
"path": "lib/model/roi_align/functions/roi_align.py",
"content": "import torch\nfrom torch.autograd import Function\nfrom .._ext import roi_align\n\n\n# TODO use save_for_backward instead\nclass RoIAlignFunction(Function):\n def __init__(self, aligned_height, aligned_width, spatial_scale):\n self.aligned_width = int(aligned_width)\n self.aligned_height = int(aligned_height)\n self.spatial_scale = float(spatial_scale)\n self.rois = None\n self.feature_size = None\n\n def forward(self, features, rois):\n self.rois = rois\n self.feature_size = features.size()\n\n batch_size, num_channels, data_height, data_width = features.size()\n num_rois = rois.size(0)\n\n output = features.new(num_rois, num_channels, self.aligned_height, self.aligned_width).zero_()\n if features.is_cuda:\n roi_align.roi_align_forward_cuda(self.aligned_height,\n self.aligned_width,\n self.spatial_scale, features,\n rois, output)\n else:\n raise NotImplementedError\n\n return output\n\n def backward(self, grad_output):\n assert(self.feature_size is not None and grad_output.is_cuda)\n\n batch_size, num_channels, data_height, data_width = self.feature_size\n\n grad_input = self.rois.new(batch_size, num_channels, data_height,\n data_width).zero_()\n roi_align.roi_align_backward_cuda(self.aligned_height,\n self.aligned_width,\n self.spatial_scale, grad_output,\n self.rois, grad_input)\n\n # print grad_input\n\n return grad_input, None\n"
},
{
"path": "lib/model/roi_align/modules/__init__.py",
"content": ""
},
{
"path": "lib/model/roi_align/modules/roi_align.py",
"content": "from torch.nn.modules.module import Module\nfrom torch.nn.functional import avg_pool2d, max_pool2d\nfrom ..functions.roi_align import RoIAlignFunction\n\n\nclass RoIAlign(Module):\n def __init__(self, aligned_height, aligned_width, spatial_scale):\n super(RoIAlign, self).__init__()\n\n self.aligned_width = int(aligned_width)\n self.aligned_height = int(aligned_height)\n self.spatial_scale = float(spatial_scale)\n\n def forward(self, features, rois):\n return RoIAlignFunction(self.aligned_height, self.aligned_width,\n self.spatial_scale)(features, rois)\n\nclass RoIAlignAvg(Module):\n def __init__(self, aligned_height, aligned_width, spatial_scale):\n super(RoIAlignAvg, self).__init__()\n\n self.aligned_width = int(aligned_width)\n self.aligned_height = int(aligned_height)\n self.spatial_scale = float(spatial_scale)\n\n def forward(self, features, rois):\n x = RoIAlignFunction(self.aligned_height+1, self.aligned_width+1,\n self.spatial_scale)(features, rois)\n return avg_pool2d(x, kernel_size=2, stride=1)\n\nclass RoIAlignMax(Module):\n def __init__(self, aligned_height, aligned_width, spatial_scale):\n super(RoIAlignMax, self).__init__()\n\n self.aligned_width = int(aligned_width)\n self.aligned_height = int(aligned_height)\n self.spatial_scale = float(spatial_scale)\n\n def forward(self, features, rois):\n x = RoIAlignFunction(self.aligned_height+1, self.aligned_width+1,\n self.spatial_scale)(features, rois)\n return max_pool2d(x, kernel_size=2, stride=1)\n"
},
{
"path": "lib/model/roi_align/src/roi_align_cuda.c",
"content": "#include \n#include \n#include \"roi_align_kernel.h\"\n\nextern THCState *state;\n\nint roi_align_forward_cuda(int aligned_height, int aligned_width, float spatial_scale,\n THCudaTensor * features, THCudaTensor * rois, THCudaTensor * output)\n{\n // Grab the input tensor\n float * data_flat = THCudaTensor_data(state, features);\n float * rois_flat = THCudaTensor_data(state, rois);\n\n float * output_flat = THCudaTensor_data(state, output);\n\n // Number of ROIs\n int num_rois = THCudaTensor_size(state, rois, 0);\n int size_rois = THCudaTensor_size(state, rois, 1);\n if (size_rois != 5)\n {\n return 0;\n }\n\n // data height\n int data_height = THCudaTensor_size(state, features, 2);\n // data width\n int data_width = THCudaTensor_size(state, features, 3);\n // Number of channels\n int num_channels = THCudaTensor_size(state, features, 1);\n\n cudaStream_t stream = THCState_getCurrentStream(state);\n\n ROIAlignForwardLaucher(\n data_flat, spatial_scale, num_rois, data_height,\n data_width, num_channels, aligned_height,\n aligned_width, rois_flat,\n output_flat, stream);\n\n return 1;\n}\n\nint roi_align_backward_cuda(int aligned_height, int aligned_width, float spatial_scale,\n THCudaTensor * top_grad, THCudaTensor * rois, THCudaTensor * bottom_grad)\n{\n // Grab the input tensor\n float * top_grad_flat = THCudaTensor_data(state, top_grad);\n float * rois_flat = THCudaTensor_data(state, rois);\n\n float * bottom_grad_flat = THCudaTensor_data(state, bottom_grad);\n\n // Number of ROIs\n int num_rois = THCudaTensor_size(state, rois, 0);\n int size_rois = THCudaTensor_size(state, rois, 1);\n if (size_rois != 5)\n {\n return 0;\n }\n\n // batch size\n int batch_size = THCudaTensor_size(state, bottom_grad, 0);\n // data height\n int data_height = THCudaTensor_size(state, bottom_grad, 2);\n // data width\n int data_width = THCudaTensor_size(state, bottom_grad, 3);\n // Number of channels\n int num_channels = THCudaTensor_size(state, bottom_grad, 1);\n\n cudaStream_t stream = THCState_getCurrentStream(state);\n ROIAlignBackwardLaucher(\n top_grad_flat, spatial_scale, batch_size, num_rois, data_height,\n data_width, num_channels, aligned_height,\n aligned_width, rois_flat,\n bottom_grad_flat, stream);\n\n return 1;\n}\n"
},
{
"path": "lib/model/roi_align/src/roi_align_cuda.h",
"content": "int roi_align_forward_cuda(int aligned_height, int aligned_width, float spatial_scale,\n THCudaTensor * features, THCudaTensor * rois, THCudaTensor * output);\n\nint roi_align_backward_cuda(int aligned_height, int aligned_width, float spatial_scale,\n THCudaTensor * top_grad, THCudaTensor * rois, THCudaTensor * bottom_grad);\n"
},
{
"path": "lib/model/roi_align/src/roi_align_kernel.cu",
"content": "#ifdef __cplusplus\nextern \"C\" {\n#endif\n\n#include \n#include \n#include \n#include \"roi_align_kernel.h\"\n\n#define CUDA_1D_KERNEL_LOOP(i, n) \\\n for (int i = blockIdx.x * blockDim.x + threadIdx.x; i < n; \\\n i += blockDim.x * gridDim.x)\n\n\n __global__ void ROIAlignForward(const int nthreads, const float* bottom_data, const float spatial_scale, const int height, const int width,\n const int channels, const int aligned_height, const int aligned_width, const float* bottom_rois, float* top_data) {\n CUDA_1D_KERNEL_LOOP(index, nthreads) {\n // (n, c, ph, pw) is an element in the aligned output\n // int n = index;\n // int pw = n % aligned_width;\n // n /= aligned_width;\n // int ph = n % aligned_height;\n // n /= aligned_height;\n // int c = n % channels;\n // n /= channels;\n\n int pw = index % aligned_width;\n int ph = (index / aligned_width) % aligned_height;\n int c = (index / aligned_width / aligned_height) % channels;\n int n = index / aligned_width / aligned_height / channels;\n\n // bottom_rois += n * 5;\n float roi_batch_ind = bottom_rois[n * 5 + 0];\n float roi_start_w = bottom_rois[n * 5 + 1] * spatial_scale;\n float roi_start_h = bottom_rois[n * 5 + 2] * spatial_scale;\n float roi_end_w = bottom_rois[n * 5 + 3] * spatial_scale;\n float roi_end_h = bottom_rois[n * 5 + 4] * spatial_scale;\n\n // Force malformed ROIs to be 1x1\n float roi_width = fmaxf(roi_end_w - roi_start_w + 1., 0.);\n float roi_height = fmaxf(roi_end_h - roi_start_h + 1., 0.);\n float bin_size_h = roi_height / (aligned_height - 1.);\n float bin_size_w = roi_width / (aligned_width - 1.);\n\n float h = (float)(ph) * bin_size_h + roi_start_h;\n float w = (float)(pw) * bin_size_w + roi_start_w;\n\n int hstart = fminf(floor(h), height - 2);\n int wstart = fminf(floor(w), width - 2);\n\n int img_start = roi_batch_ind * channels * height * width;\n\n // bilinear interpolation\n if (h < 0 || h >= height || w < 0 || w >= width) {\n top_data[index] = 0.;\n } else {\n float h_ratio = h - (float)(hstart);\n float w_ratio = w - (float)(wstart);\n int upleft = img_start + (c * height + hstart) * width + wstart;\n int upright = upleft + 1;\n int downleft = upleft + width;\n int downright = downleft + 1;\n\n top_data[index] = bottom_data[upleft] * (1. - h_ratio) * (1. - w_ratio)\n + bottom_data[upright] * (1. - h_ratio) * w_ratio\n + bottom_data[downleft] * h_ratio * (1. - w_ratio)\n + bottom_data[downright] * h_ratio * w_ratio;\n }\n }\n }\n\n\n int ROIAlignForwardLaucher(const float* bottom_data, const float spatial_scale, const int num_rois, const int height, const int width,\n const int channels, const int aligned_height, const int aligned_width, const float* bottom_rois, float* top_data, cudaStream_t stream) {\n const int kThreadsPerBlock = 1024;\n const int output_size = num_rois * aligned_height * aligned_width * channels;\n cudaError_t err;\n\n\n ROIAlignForward<<<(output_size + kThreadsPerBlock - 1) / kThreadsPerBlock, kThreadsPerBlock, 0, stream>>>(\n output_size, bottom_data, spatial_scale, height, width, channels,\n aligned_height, aligned_width, bottom_rois, top_data);\n\n err = cudaGetLastError();\n if(cudaSuccess != err) {\n fprintf( stderr, \"cudaCheckError() failed : %s\\n\", cudaGetErrorString( err ) );\n exit( -1 );\n }\n\n return 1;\n }\n\n\n __global__ void ROIAlignBackward(const int nthreads, const float* top_diff, const float spatial_scale, const int height, const int width,\n const int channels, const int aligned_height, const int aligned_width, float* bottom_diff, const float* bottom_rois) {\n CUDA_1D_KERNEL_LOOP(index, nthreads) {\n\n // (n, c, ph, pw) is an element in the aligned output\n int pw = index % aligned_width;\n int ph = (index / aligned_width) % aligned_height;\n int c = (index / aligned_width / aligned_height) % channels;\n int n = index / aligned_width / aligned_height / channels;\n\n float roi_batch_ind = bottom_rois[n * 5 + 0];\n float roi_start_w = bottom_rois[n * 5 + 1] * spatial_scale;\n float roi_start_h = bottom_rois[n * 5 + 2] * spatial_scale;\n float roi_end_w = bottom_rois[n * 5 + 3] * spatial_scale;\n float roi_end_h = bottom_rois[n * 5 + 4] * spatial_scale;\n /* int roi_start_w = round(bottom_rois[1] * spatial_scale); */\n /* int roi_start_h = round(bottom_rois[2] * spatial_scale); */\n /* int roi_end_w = round(bottom_rois[3] * spatial_scale); */\n /* int roi_end_h = round(bottom_rois[4] * spatial_scale); */\n\n // Force malformed ROIs to be 1x1\n float roi_width = fmaxf(roi_end_w - roi_start_w + 1., 0.);\n float roi_height = fmaxf(roi_end_h - roi_start_h + 1., 0.);\n float bin_size_h = roi_height / (aligned_height - 1.);\n float bin_size_w = roi_width / (aligned_width - 1.);\n\n float h = (float)(ph) * bin_size_h + roi_start_h;\n float w = (float)(pw) * bin_size_w + roi_start_w;\n\n int hstart = fminf(floor(h), height - 2);\n int wstart = fminf(floor(w), width - 2);\n\n int img_start = roi_batch_ind * channels * height * width;\n\n // bilinear interpolation\n if (!(h < 0 || h >= height || w < 0 || w >= width)) {\n float h_ratio = h - (float)(hstart);\n float w_ratio = w - (float)(wstart);\n int upleft = img_start + (c * height + hstart) * width + wstart;\n int upright = upleft + 1;\n int downleft = upleft + width;\n int downright = downleft + 1;\n\n atomicAdd(bottom_diff + upleft, top_diff[index] * (1. - h_ratio) * (1 - w_ratio));\n atomicAdd(bottom_diff + upright, top_diff[index] * (1. - h_ratio) * w_ratio);\n atomicAdd(bottom_diff + downleft, top_diff[index] * h_ratio * (1 - w_ratio));\n atomicAdd(bottom_diff + downright, top_diff[index] * h_ratio * w_ratio);\n }\n }\n }\n\n int ROIAlignBackwardLaucher(const float* top_diff, const float spatial_scale, const int batch_size, const int num_rois, const int height, const int width,\n const int channels, const int aligned_height, const int aligned_width, const float* bottom_rois, float* bottom_diff, cudaStream_t stream) {\n const int kThreadsPerBlock = 1024;\n const int output_size = num_rois * aligned_height * aligned_width * channels;\n cudaError_t err;\n\n ROIAlignBackward<<<(output_size + kThreadsPerBlock - 1) / kThreadsPerBlock, kThreadsPerBlock, 0, stream>>>(\n output_size, top_diff, spatial_scale, height, width, channels,\n aligned_height, aligned_width, bottom_diff, bottom_rois);\n\n err = cudaGetLastError();\n if(cudaSuccess != err) {\n fprintf( stderr, \"cudaCheckError() failed : %s\\n\", cudaGetErrorString( err ) );\n exit( -1 );\n }\n\n return 1;\n }\n\n\n#ifdef __cplusplus\n}\n#endif\n"
},
{
"path": "lib/model/roi_align/src/roi_align_kernel.h",
"content": "#ifndef _ROI_ALIGN_KERNEL\n#define _ROI_ALIGN_KERNEL\n\n#ifdef __cplusplus\nextern \"C\" {\n#endif\n\n__global__ void ROIAlignForward(const int nthreads, const float* bottom_data,\n const float spatial_scale, const int height, const int width,\n const int channels, const int aligned_height, const int aligned_width,\n const float* bottom_rois, float* top_data);\n\nint ROIAlignForwardLaucher(\n const float* bottom_data, const float spatial_scale, const int num_rois, const int height,\n const int width, const int channels, const int aligned_height,\n const int aligned_width, const float* bottom_rois,\n float* top_data, cudaStream_t stream);\n\n__global__ void ROIAlignBackward(const int nthreads, const float* top_diff,\n const float spatial_scale, const int height, const int width,\n const int channels, const int aligned_height, const int aligned_width,\n float* bottom_diff, const float* bottom_rois);\n\nint ROIAlignBackwardLaucher(const float* top_diff, const float spatial_scale, const int batch_size, const int num_rois,\n const int height, const int width, const int channels, const int aligned_height,\n const int aligned_width, const float* bottom_rois,\n float* bottom_diff, cudaStream_t stream);\n\n#ifdef __cplusplus\n}\n#endif\n\n#endif\n\n"
},
{
"path": "lib/model/roi_crop/__init__.py",
"content": ""
},
{
"path": "lib/model/roi_crop/_ext/__init__.py",
"content": ""
},
{
"path": "lib/model/roi_crop/_ext/roi_crop/__init__.py",
"content": "\nfrom torch.utils.ffi import _wrap_function\nfrom ._roi_crop import lib as _lib, ffi as _ffi\n\n__all__ = []\ndef _import_symbols(locals):\n for symbol in dir(_lib):\n fn = getattr(_lib, symbol)\n if callable(fn):\n locals[symbol] = _wrap_function(fn, _ffi)\n else:\n locals[symbol] = fn\n __all__.append(symbol)\n\n_import_symbols(locals())\n"
},
{
"path": "lib/model/roi_crop/build.py",
"content": "from __future__ import print_function\nimport os\nimport torch\nfrom torch.utils.ffi import create_extension\n\n#this_file = os.path.dirname(__file__)\n\nsources = ['src/roi_crop.c']\nheaders = ['src/roi_crop.h']\ndefines = []\nwith_cuda = False\n\nif torch.cuda.is_available():\n print('Including CUDA code.')\n sources += ['src/roi_crop_cuda.c']\n headers += ['src/roi_crop_cuda.h']\n defines += [('WITH_CUDA', None)]\n with_cuda = True\n\nthis_file = os.path.dirname(os.path.realpath(__file__))\nprint(this_file)\nextra_objects = ['src/roi_crop_cuda_kernel.cu.o']\nextra_objects = [os.path.join(this_file, fname) for fname in extra_objects]\n\nffi = create_extension(\n '_ext.roi_crop',\n headers=headers,\n sources=sources,\n define_macros=defines,\n relative_to=__file__,\n with_cuda=with_cuda,\n extra_objects=extra_objects\n)\n\nif __name__ == '__main__':\n ffi.build()\n"
},
{
"path": "lib/model/roi_crop/functions/__init__.py",
"content": ""
},
{
"path": "lib/model/roi_crop/functions/crop_resize.py",
"content": "# functions/add.py\nimport torch\nfrom torch.autograd import Function\nfrom .._ext import roi_crop\nfrom cffi import FFI\nffi = FFI()\n\nclass RoICropFunction(Function):\n def forward(self, input1, input2):\n self.input1 = input1\n self.input2 = input2\n self.device_c = ffi.new(\"int *\")\n output = torch.zeros(input2.size()[0], input1.size()[1], input2.size()[1], input2.size()[2])\n #print('decice %d' % torch.cuda.current_device())\n if input1.is_cuda:\n self.device = torch.cuda.current_device()\n else:\n self.device = -1\n self.device_c[0] = self.device\n if not input1.is_cuda:\n roi_crop.BilinearSamplerBHWD_updateOutput(input1, input2, output)\n else:\n output = output.cuda(self.device)\n roi_crop.BilinearSamplerBHWD_updateOutput_cuda(input1, input2, output)\n return output\n\n def backward(self, grad_output):\n grad_input1 = torch.zeros(self.input1.size())\n grad_input2 = torch.zeros(self.input2.size())\n #print('backward decice %d' % self.device)\n if not grad_output.is_cuda:\n roi_crop.BilinearSamplerBHWD_updateGradInput(self.input1, self.input2, grad_input1, grad_input2, grad_output)\n else:\n grad_input1 = grad_input1.cuda(self.device)\n grad_input2 = grad_input2.cuda(self.device)\n roi_crop.BilinearSamplerBHWD_updateGradInput_cuda(self.input1, self.input2, grad_input1, grad_input2, grad_output)\n return grad_input1, grad_input2\n"
},
{
"path": "lib/model/roi_crop/functions/gridgen.py",
"content": "# functions/add.py\nimport torch\nfrom torch.autograd import Function\nimport numpy as np\n\n\nclass AffineGridGenFunction(Function):\n def __init__(self, height, width,lr=1):\n super(AffineGridGenFunction, self).__init__()\n self.lr = lr\n self.height, self.width = height, width\n self.grid = np.zeros( [self.height, self.width, 3], dtype=np.float32)\n self.grid[:,:,0] = np.expand_dims(np.repeat(np.expand_dims(np.arange(-1, 1, 2.0/(self.height)), 0), repeats = self.width, axis = 0).T, 0)\n self.grid[:,:,1] = np.expand_dims(np.repeat(np.expand_dims(np.arange(-1, 1, 2.0/(self.width)), 0), repeats = self.height, axis = 0), 0)\n # self.grid[:,:,0] = np.expand_dims(np.repeat(np.expand_dims(np.arange(-1, 1, 2.0/(self.height - 1)), 0), repeats = self.width, axis = 0).T, 0)\n # self.grid[:,:,1] = np.expand_dims(np.repeat(np.expand_dims(np.arange(-1, 1, 2.0/(self.width - 1)), 0), repeats = self.height, axis = 0), 0)\n self.grid[:,:,2] = np.ones([self.height, width])\n self.grid = torch.from_numpy(self.grid.astype(np.float32))\n #print(self.grid)\n\n def forward(self, input1):\n self.input1 = input1\n output = input1.new(torch.Size([input1.size(0)]) + self.grid.size()).zero_()\n self.batchgrid = input1.new(torch.Size([input1.size(0)]) + self.grid.size()).zero_()\n for i in range(input1.size(0)):\n self.batchgrid[i] = self.grid.astype(self.batchgrid[i])\n\n # if input1.is_cuda:\n # self.batchgrid = self.batchgrid.cuda()\n # output = output.cuda()\n\n for i in range(input1.size(0)):\n output = torch.bmm(self.batchgrid.view(-1, self.height*self.width, 3), torch.transpose(input1, 1, 2)).view(-1, self.height, self.width, 2)\n\n return output\n\n def backward(self, grad_output):\n\n grad_input1 = self.input1.new(self.input1.size()).zero_()\n\n # if grad_output.is_cuda:\n # self.batchgrid = self.batchgrid.cuda()\n # grad_input1 = grad_input1.cuda()\n\n grad_input1 = torch.baddbmm(grad_input1, torch.transpose(grad_output.view(-1, self.height*self.width, 2), 1,2), self.batchgrid.view(-1, self.height*self.width, 3))\n return grad_input1\n"
},
{
"path": "lib/model/roi_crop/functions/roi_crop.py",
"content": "# functions/add.py\nimport torch\nfrom torch.autograd import Function\nfrom .._ext import roi_crop\nimport pdb\n\nclass RoICropFunction(Function):\n def forward(self, input1, input2):\n self.input1 = input1.clone()\n self.input2 = input2.clone()\n output = input2.new(input2.size()[0], input1.size()[1], input2.size()[1], input2.size()[2]).zero_()\n assert output.get_device() == input1.get_device(), \"output and input1 must on the same device\"\n assert output.get_device() == input2.get_device(), \"output and input2 must on the same device\"\n roi_crop.BilinearSamplerBHWD_updateOutput_cuda(input1, input2, output)\n return output\n\n def backward(self, grad_output):\n grad_input1 = self.input1.new(self.input1.size()).zero_()\n grad_input2 = self.input2.new(self.input2.size()).zero_()\n roi_crop.BilinearSamplerBHWD_updateGradInput_cuda(self.input1, self.input2, grad_input1, grad_input2, grad_output)\n return grad_input1, grad_input2\n"
},
{
"path": "lib/model/roi_crop/modules/__init__.py",
"content": ""
},
{
"path": "lib/model/roi_crop/modules/gridgen.py",
"content": "from torch.nn.modules.module import Module\nimport torch\nfrom torch.autograd import Variable\nimport numpy as np\nfrom ..functions.gridgen import AffineGridGenFunction\n\nimport pyximport\npyximport.install(setup_args={\"include_dirs\":np.get_include()},\n reload_support=True)\n\n\nclass _AffineGridGen(Module):\n def __init__(self, height, width, lr = 1, aux_loss = False):\n super(_AffineGridGen, self).__init__()\n self.height, self.width = height, width\n self.aux_loss = aux_loss\n self.f = AffineGridGenFunction(self.height, self.width, lr=lr)\n self.lr = lr\n def forward(self, input):\n # if not self.aux_loss:\n return self.f(input)\n # else:\n # identity = torch.from_numpy(np.array([[1,0,0], [0,1,0]], dtype=np.float32))\n # batch_identity = torch.zeros([input.size(0), 2,3])\n # for i in range(input.size(0)):\n # batch_identity[i] = identity\n # batch_identity = Variable(batch_identity)\n # loss = torch.mul(input - batch_identity, input - batch_identity)\n # loss = torch.sum(loss,1)\n # loss = torch.sum(loss,2)\n\n # return self.f(input), loss.view(-1,1)\n\nclass CylinderGridGen(Module):\n def __init__(self, height, width, lr = 1, aux_loss = False):\n super(CylinderGridGen, self).__init__()\n self.height, self.width = height, width\n self.aux_loss = aux_loss\n self.f = CylinderGridGenFunction(self.height, self.width, lr=lr)\n self.lr = lr\n def forward(self, input):\n\n if not self.aux_loss:\n return self.f(input)\n else:\n return self.f(input), torch.mul(input, input).view(-1,1)\n\n\nclass AffineGridGenV2(Module):\n def __init__(self, height, width, lr = 1, aux_loss = False):\n super(AffineGridGenV2, self).__init__()\n self.height, self.width = height, width\n self.aux_loss = aux_loss\n self.lr = lr\n\n self.grid = np.zeros( [self.height, self.width, 3], dtype=np.float32)\n self.grid[:,:,0] = np.expand_dims(np.repeat(np.expand_dims(np.arange(-1, 1, 2.0/self.height), 0), repeats = self.width, axis = 0).T, 0)\n self.grid[:,:,1] = np.expand_dims(np.repeat(np.expand_dims(np.arange(-1, 1, 2.0/self.width), 0), repeats = self.height, axis = 0), 0)\n self.grid[:,:,2] = np.ones([self.height, width])\n self.grid = torch.from_numpy(self.grid.astype(np.float32))\n\n\n def forward(self, input1):\n self.batchgrid = torch.zeros(torch.Size([input1.size(0)]) + self.grid.size())\n\n for i in range(input1.size(0)):\n self.batchgrid[i] = self.grid\n self.batchgrid = Variable(self.batchgrid)\n\n if input1.is_cuda:\n self.batchgrid = self.batchgrid.cuda()\n\n output = torch.bmm(self.batchgrid.view(-1, self.height*self.width, 3), torch.transpose(input1, 1, 2)).view(-1, self.height, self.width, 2)\n\n return output\n\n\nclass CylinderGridGenV2(Module):\n def __init__(self, height, width, lr = 1):\n super(CylinderGridGenV2, self).__init__()\n self.height, self.width = height, width\n self.lr = lr\n self.grid = np.zeros( [self.height, self.width, 3], dtype=np.float32)\n self.grid[:,:,0] = np.expand_dims(np.repeat(np.expand_dims(np.arange(-1, 1, 2.0/self.height), 0), repeats = self.width, axis = 0).T, 0)\n self.grid[:,:,1] = np.expand_dims(np.repeat(np.expand_dims(np.arange(-1, 1, 2.0/self.width), 0), repeats = self.height, axis = 0), 0)\n self.grid[:,:,2] = np.ones([self.height, width])\n self.grid = torch.from_numpy(self.grid.astype(np.float32))\n def forward(self, input):\n self.batchgrid = torch.zeros(torch.Size([input.size(0)]) + self.grid.size() )\n #print(self.batchgrid.size())\n for i in range(input.size(0)):\n self.batchgrid[i,:,:,:] = self.grid\n self.batchgrid = Variable(self.batchgrid)\n\n #print(self.batchgrid.size())\n\n input_u = input.view(-1,1,1,1).repeat(1,self.height, self.width,1)\n #print(input_u.requires_grad, self.batchgrid)\n\n output0 = self.batchgrid[:,:,:,0:1]\n output1 = torch.atan(torch.tan(np.pi/2.0*(self.batchgrid[:,:,:,1:2] + self.batchgrid[:,:,:,2:] * input_u[:,:,:,:]))) /(np.pi/2)\n #print(output0.size(), output1.size())\n\n output = torch.cat([output0, output1], 3)\n return output\n\n\nclass DenseAffineGridGen(Module):\n def __init__(self, height, width, lr = 1, aux_loss = False):\n super(DenseAffineGridGen, self).__init__()\n self.height, self.width = height, width\n self.aux_loss = aux_loss\n self.lr = lr\n\n self.grid = np.zeros( [self.height, self.width, 3], dtype=np.float32)\n self.grid[:,:,0] = np.expand_dims(np.repeat(np.expand_dims(np.arange(-1, 1, 2.0/self.height), 0), repeats = self.width, axis = 0).T, 0)\n self.grid[:,:,1] = np.expand_dims(np.repeat(np.expand_dims(np.arange(-1, 1, 2.0/self.width), 0), repeats = self.height, axis = 0), 0)\n self.grid[:,:,2] = np.ones([self.height, width])\n self.grid = torch.from_numpy(self.grid.astype(np.float32))\n\n\n def forward(self, input1):\n self.batchgrid = torch.zeros(torch.Size([input1.size(0)]) + self.grid.size())\n\n for i in range(input1.size(0)):\n self.batchgrid[i] = self.grid\n\n self.batchgrid = Variable(self.batchgrid)\n #print self.batchgrid, input1[:,:,:,0:3]\n #print self.batchgrid, input1[:,:,:,4:6]\n x = torch.mul(self.batchgrid, input1[:,:,:,0:3])\n y = torch.mul(self.batchgrid, input1[:,:,:,3:6])\n\n output = torch.cat([torch.sum(x,3),torch.sum(y,3)], 3)\n return output\n\n\n\n\nclass DenseAffine3DGridGen(Module):\n def __init__(self, height, width, lr = 1, aux_loss = False):\n super(DenseAffine3DGridGen, self).__init__()\n self.height, self.width = height, width\n self.aux_loss = aux_loss\n self.lr = lr\n\n self.grid = np.zeros( [self.height, self.width, 3], dtype=np.float32)\n self.grid[:,:,0] = np.expand_dims(np.repeat(np.expand_dims(np.arange(-1, 1, 2.0/self.height), 0), repeats = self.width, axis = 0).T, 0)\n self.grid[:,:,1] = np.expand_dims(np.repeat(np.expand_dims(np.arange(-1, 1, 2.0/self.width), 0), repeats = self.height, axis = 0), 0)\n self.grid[:,:,2] = np.ones([self.height, width])\n self.grid = torch.from_numpy(self.grid.astype(np.float32))\n\n self.theta = self.grid[:,:,0] * np.pi/2 + np.pi/2\n self.phi = self.grid[:,:,1] * np.pi\n\n self.x = torch.sin(self.theta) * torch.cos(self.phi)\n self.y = torch.sin(self.theta) * torch.sin(self.phi)\n self.z = torch.cos(self.theta)\n\n self.grid3d = torch.from_numpy(np.zeros( [self.height, self.width, 4], dtype=np.float32))\n\n self.grid3d[:,:,0] = self.x\n self.grid3d[:,:,1] = self.y\n self.grid3d[:,:,2] = self.z\n self.grid3d[:,:,3] = self.grid[:,:,2]\n\n\n def forward(self, input1):\n self.batchgrid3d = torch.zeros(torch.Size([input1.size(0)]) + self.grid3d.size())\n\n for i in range(input1.size(0)):\n self.batchgrid3d[i] = self.grid3d\n\n self.batchgrid3d = Variable(self.batchgrid3d)\n #print(self.batchgrid3d)\n\n x = torch.sum(torch.mul(self.batchgrid3d, input1[:,:,:,0:4]), 3)\n y = torch.sum(torch.mul(self.batchgrid3d, input1[:,:,:,4:8]), 3)\n z = torch.sum(torch.mul(self.batchgrid3d, input1[:,:,:,8:]), 3)\n #print(x)\n r = torch.sqrt(x**2 + y**2 + z**2) + 1e-5\n\n #print(r)\n theta = torch.acos(z/r)/(np.pi/2) - 1\n #phi = torch.atan(y/x)\n phi = torch.atan(y/(x + 1e-5)) + np.pi * x.lt(0).type(torch.FloatTensor) * (y.ge(0).type(torch.FloatTensor) - y.lt(0).type(torch.FloatTensor))\n phi = phi/np.pi\n\n\n output = torch.cat([theta,phi], 3)\n\n return output\n\n\n\n\n\nclass DenseAffine3DGridGen_rotate(Module):\n def __init__(self, height, width, lr = 1, aux_loss = False):\n super(DenseAffine3DGridGen_rotate, self).__init__()\n self.height, self.width = height, width\n self.aux_loss = aux_loss\n self.lr = lr\n\n self.grid = np.zeros( [self.height, self.width, 3], dtype=np.float32)\n self.grid[:,:,0] = np.expand_dims(np.repeat(np.expand_dims(np.arange(-1, 1, 2.0/self.height), 0), repeats = self.width, axis = 0).T, 0)\n self.grid[:,:,1] = np.expand_dims(np.repeat(np.expand_dims(np.arange(-1, 1, 2.0/self.width), 0), repeats = self.height, axis = 0), 0)\n self.grid[:,:,2] = np.ones([self.height, width])\n self.grid = torch.from_numpy(self.grid.astype(np.float32))\n\n self.theta = self.grid[:,:,0] * np.pi/2 + np.pi/2\n self.phi = self.grid[:,:,1] * np.pi\n\n self.x = torch.sin(self.theta) * torch.cos(self.phi)\n self.y = torch.sin(self.theta) * torch.sin(self.phi)\n self.z = torch.cos(self.theta)\n\n self.grid3d = torch.from_numpy(np.zeros( [self.height, self.width, 4], dtype=np.float32))\n\n self.grid3d[:,:,0] = self.x\n self.grid3d[:,:,1] = self.y\n self.grid3d[:,:,2] = self.z\n self.grid3d[:,:,3] = self.grid[:,:,2]\n\n\n def forward(self, input1, input2):\n self.batchgrid3d = torch.zeros(torch.Size([input1.size(0)]) + self.grid3d.size())\n\n for i in range(input1.size(0)):\n self.batchgrid3d[i] = self.grid3d\n\n self.batchgrid3d = Variable(self.batchgrid3d)\n\n self.batchgrid = torch.zeros(torch.Size([input1.size(0)]) + self.grid.size())\n\n for i in range(input1.size(0)):\n self.batchgrid[i] = self.grid\n\n self.batchgrid = Variable(self.batchgrid)\n\n #print(self.batchgrid3d)\n\n x = torch.sum(torch.mul(self.batchgrid3d, input1[:,:,:,0:4]), 3)\n y = torch.sum(torch.mul(self.batchgrid3d, input1[:,:,:,4:8]), 3)\n z = torch.sum(torch.mul(self.batchgrid3d, input1[:,:,:,8:]), 3)\n #print(x)\n r = torch.sqrt(x**2 + y**2 + z**2) + 1e-5\n\n #print(r)\n theta = torch.acos(z/r)/(np.pi/2) - 1\n #phi = torch.atan(y/x)\n phi = torch.atan(y/(x + 1e-5)) + np.pi * x.lt(0).type(torch.FloatTensor) * (y.ge(0).type(torch.FloatTensor) - y.lt(0).type(torch.FloatTensor))\n phi = phi/np.pi\n\n input_u = input2.view(-1,1,1,1).repeat(1,self.height, self.width,1)\n\n output = torch.cat([theta,phi], 3)\n\n output1 = torch.atan(torch.tan(np.pi/2.0*(output[:,:,:,1:2] + self.batchgrid[:,:,:,2:] * input_u[:,:,:,:]))) /(np.pi/2)\n output2 = torch.cat([output[:,:,:,0:1], output1], 3)\n\n return output2\n\n\nclass Depth3DGridGen(Module):\n def __init__(self, height, width, lr = 1, aux_loss = False):\n super(Depth3DGridGen, self).__init__()\n self.height, self.width = height, width\n self.aux_loss = aux_loss\n self.lr = lr\n\n self.grid = np.zeros( [self.height, self.width, 3], dtype=np.float32)\n self.grid[:,:,0] = np.expand_dims(np.repeat(np.expand_dims(np.arange(-1, 1, 2.0/self.height), 0), repeats = self.width, axis = 0).T, 0)\n self.grid[:,:,1] = np.expand_dims(np.repeat(np.expand_dims(np.arange(-1, 1, 2.0/self.width), 0), repeats = self.height, axis = 0), 0)\n self.grid[:,:,2] = np.ones([self.height, width])\n self.grid = torch.from_numpy(self.grid.astype(np.float32))\n\n self.theta = self.grid[:,:,0] * np.pi/2 + np.pi/2\n self.phi = self.grid[:,:,1] * np.pi\n\n self.x = torch.sin(self.theta) * torch.cos(self.phi)\n self.y = torch.sin(self.theta) * torch.sin(self.phi)\n self.z = torch.cos(self.theta)\n\n self.grid3d = torch.from_numpy(np.zeros( [self.height, self.width, 4], dtype=np.float32))\n\n self.grid3d[:,:,0] = self.x\n self.grid3d[:,:,1] = self.y\n self.grid3d[:,:,2] = self.z\n self.grid3d[:,:,3] = self.grid[:,:,2]\n\n\n def forward(self, depth, trans0, trans1, rotate):\n self.batchgrid3d = torch.zeros(torch.Size([depth.size(0)]) + self.grid3d.size())\n\n for i in range(depth.size(0)):\n self.batchgrid3d[i] = self.grid3d\n\n self.batchgrid3d = Variable(self.batchgrid3d)\n\n self.batchgrid = torch.zeros(torch.Size([depth.size(0)]) + self.grid.size())\n\n for i in range(depth.size(0)):\n self.batchgrid[i] = self.grid\n\n self.batchgrid = Variable(self.batchgrid)\n\n x = self.batchgrid3d[:,:,:,0:1] * depth + trans0.view(-1,1,1,1).repeat(1, self.height, self.width, 1)\n\n y = self.batchgrid3d[:,:,:,1:2] * depth + trans1.view(-1,1,1,1).repeat(1, self.height, self.width, 1)\n z = self.batchgrid3d[:,:,:,2:3] * depth\n #print(x.size(), y.size(), z.size())\n r = torch.sqrt(x**2 + y**2 + z**2) + 1e-5\n\n #print(r)\n theta = torch.acos(z/r)/(np.pi/2) - 1\n #phi = torch.atan(y/x)\n phi = torch.atan(y/(x + 1e-5)) + np.pi * x.lt(0).type(torch.FloatTensor) * (y.ge(0).type(torch.FloatTensor) - y.lt(0).type(torch.FloatTensor))\n phi = phi/np.pi\n\n #print(theta.size(), phi.size())\n\n\n input_u = rotate.view(-1,1,1,1).repeat(1,self.height, self.width,1)\n\n output = torch.cat([theta,phi], 3)\n #print(output.size())\n\n output1 = torch.atan(torch.tan(np.pi/2.0*(output[:,:,:,1:2] + self.batchgrid[:,:,:,2:] * input_u[:,:,:,:]))) /(np.pi/2)\n output2 = torch.cat([output[:,:,:,0:1], output1], 3)\n\n return output2\n\n\n\n\n\nclass Depth3DGridGen_with_mask(Module):\n def __init__(self, height, width, lr = 1, aux_loss = False, ray_tracing = False):\n super(Depth3DGridGen_with_mask, self).__init__()\n self.height, self.width = height, width\n self.aux_loss = aux_loss\n self.lr = lr\n self.ray_tracing = ray_tracing\n\n self.grid = np.zeros( [self.height, self.width, 3], dtype=np.float32)\n self.grid[:,:,0] = np.expand_dims(np.repeat(np.expand_dims(np.arange(-1, 1, 2.0/self.height), 0), repeats = self.width, axis = 0).T, 0)\n self.grid[:,:,1] = np.expand_dims(np.repeat(np.expand_dims(np.arange(-1, 1, 2.0/self.width), 0), repeats = self.height, axis = 0), 0)\n self.grid[:,:,2] = np.ones([self.height, width])\n self.grid = torch.from_numpy(self.grid.astype(np.float32))\n\n self.theta = self.grid[:,:,0] * np.pi/2 + np.pi/2\n self.phi = self.grid[:,:,1] * np.pi\n\n self.x = torch.sin(self.theta) * torch.cos(self.phi)\n self.y = torch.sin(self.theta) * torch.sin(self.phi)\n self.z = torch.cos(self.theta)\n\n self.grid3d = torch.from_numpy(np.zeros( [self.height, self.width, 4], dtype=np.float32))\n\n self.grid3d[:,:,0] = self.x\n self.grid3d[:,:,1] = self.y\n self.grid3d[:,:,2] = self.z\n self.grid3d[:,:,3] = self.grid[:,:,2]\n\n\n def forward(self, depth, trans0, trans1, rotate):\n self.batchgrid3d = torch.zeros(torch.Size([depth.size(0)]) + self.grid3d.size())\n\n for i in range(depth.size(0)):\n self.batchgrid3d[i] = self.grid3d\n\n self.batchgrid3d = Variable(self.batchgrid3d)\n\n self.batchgrid = torch.zeros(torch.Size([depth.size(0)]) + self.grid.size())\n\n for i in range(depth.size(0)):\n self.batchgrid[i] = self.grid\n\n self.batchgrid = Variable(self.batchgrid)\n\n if depth.is_cuda:\n self.batchgrid = self.batchgrid.cuda()\n self.batchgrid3d = self.batchgrid3d.cuda()\n\n\n x_ = self.batchgrid3d[:,:,:,0:1] * depth + trans0.view(-1,1,1,1).repeat(1, self.height, self.width, 1)\n\n y_ = self.batchgrid3d[:,:,:,1:2] * depth + trans1.view(-1,1,1,1).repeat(1, self.height, self.width, 1)\n z = self.batchgrid3d[:,:,:,2:3] * depth\n #print(x.size(), y.size(), z.size())\n\n rotate_z = rotate.view(-1,1,1,1).repeat(1,self.height, self.width,1) * np.pi\n\n x = x_ * torch.cos(rotate_z) - y_ * torch.sin(rotate_z)\n y = x_ * torch.sin(rotate_z) + y_ * torch.cos(rotate_z)\n\n\n r = torch.sqrt(x**2 + y**2 + z**2) + 1e-5\n\n #print(r)\n theta = torch.acos(z/r)/(np.pi/2) - 1\n #phi = torch.atan(y/x)\n\n if depth.is_cuda:\n phi = torch.atan(y/(x + 1e-5)) + np.pi * x.lt(0).type(torch.cuda.FloatTensor) * (y.ge(0).type(torch.cuda.FloatTensor) - y.lt(0).type(torch.cuda.FloatTensor))\n else:\n phi = torch.atan(y/(x + 1e-5)) + np.pi * x.lt(0).type(torch.FloatTensor) * (y.ge(0).type(torch.FloatTensor) - y.lt(0).type(torch.FloatTensor))\n\n\n phi = phi/np.pi\n\n output = torch.cat([theta,phi], 3)\n return output\n"
},
{
"path": "lib/model/roi_crop/modules/roi_crop.py",
"content": "from torch.nn.modules.module import Module\nfrom ..functions.roi_crop import RoICropFunction\n\nclass _RoICrop(Module):\n def __init__(self, layout = 'BHWD'):\n super(_RoICrop, self).__init__()\n def forward(self, input1, input2):\n return RoICropFunction()(input1, input2)\n"
},
{
"path": "lib/model/roi_crop/src/roi_crop.c",
"content": "#include \n#include \n#include \n\n#define real float\n\nint BilinearSamplerBHWD_updateOutput(THFloatTensor *inputImages, THFloatTensor *grids, THFloatTensor *output)\n{\n\n int batchsize = inputImages->size[0];\n int inputImages_height = inputImages->size[1];\n int inputImages_width = inputImages->size[2];\n int output_height = output->size[1];\n int output_width = output->size[2];\n int inputImages_channels = inputImages->size[3];\n\n int output_strideBatch = output->stride[0];\n int output_strideHeight = output->stride[1];\n int output_strideWidth = output->stride[2];\n\n int inputImages_strideBatch = inputImages->stride[0];\n int inputImages_strideHeight = inputImages->stride[1];\n int inputImages_strideWidth = inputImages->stride[2];\n\n int grids_strideBatch = grids->stride[0];\n int grids_strideHeight = grids->stride[1];\n int grids_strideWidth = grids->stride[2];\n\n\n real *inputImages_data, *output_data, *grids_data;\n inputImages_data = THFloatTensor_data(inputImages);\n output_data = THFloatTensor_data(output);\n grids_data = THFloatTensor_data(grids);\n\n int b, yOut, xOut;\n\n for(b=0; b < batchsize; b++)\n {\n for(yOut=0; yOut < output_height; yOut++)\n {\n for(xOut=0; xOut < output_width; xOut++)\n {\n //read the grid\n real yf = grids_data[b*grids_strideBatch + yOut*grids_strideHeight + xOut*grids_strideWidth];\n real xf = grids_data[b*grids_strideBatch + yOut*grids_strideHeight + xOut*grids_strideWidth + 1];\n\n // get the weights for interpolation\n int yInTopLeft, xInTopLeft;\n real yWeightTopLeft, xWeightTopLeft;\n\n real xcoord = (xf + 1) * (inputImages_width - 1) / 2;\n xInTopLeft = floor(xcoord);\n xWeightTopLeft = 1 - (xcoord - xInTopLeft);\n\n real ycoord = (yf + 1) * (inputImages_height - 1) / 2;\n yInTopLeft = floor(ycoord);\n yWeightTopLeft = 1 - (ycoord - yInTopLeft);\n\n\n\n const int outAddress = output_strideBatch * b + output_strideHeight * yOut + output_strideWidth * xOut;\n const int inTopLeftAddress = inputImages_strideBatch * b + inputImages_strideHeight * yInTopLeft + inputImages_strideWidth * xInTopLeft;\n const int inTopRightAddress = inTopLeftAddress + inputImages_strideWidth;\n const int inBottomLeftAddress = inTopLeftAddress + inputImages_strideHeight;\n const int inBottomRightAddress = inBottomLeftAddress + inputImages_strideWidth;\n\n real v=0;\n real inTopLeft=0;\n real inTopRight=0;\n real inBottomLeft=0;\n real inBottomRight=0;\n\n // we are careful with the boundaries\n bool topLeftIsIn = xInTopLeft >= 0 && xInTopLeft <= inputImages_width-1 && yInTopLeft >= 0 && yInTopLeft <= inputImages_height-1;\n bool topRightIsIn = xInTopLeft+1 >= 0 && xInTopLeft+1 <= inputImages_width-1 && yInTopLeft >= 0 && yInTopLeft <= inputImages_height-1;\n bool bottomLeftIsIn = xInTopLeft >= 0 && xInTopLeft <= inputImages_width-1 && yInTopLeft+1 >= 0 && yInTopLeft+1 <= inputImages_height-1;\n bool bottomRightIsIn = xInTopLeft+1 >= 0 && xInTopLeft+1 <= inputImages_width-1 && yInTopLeft+1 >= 0 && yInTopLeft+1 <= inputImages_height-1;\n\n int t;\n // interpolation happens here\n for(t=0; tsize[0];\n int inputImages_height = inputImages->size[1];\n int inputImages_width = inputImages->size[2];\n int gradOutput_height = gradOutput->size[1];\n int gradOutput_width = gradOutput->size[2];\n int inputImages_channels = inputImages->size[3];\n\n int gradOutput_strideBatch = gradOutput->stride[0];\n int gradOutput_strideHeight = gradOutput->stride[1];\n int gradOutput_strideWidth = gradOutput->stride[2];\n\n int inputImages_strideBatch = inputImages->stride[0];\n int inputImages_strideHeight = inputImages->stride[1];\n int inputImages_strideWidth = inputImages->stride[2];\n\n int gradInputImages_strideBatch = gradInputImages->stride[0];\n int gradInputImages_strideHeight = gradInputImages->stride[1];\n int gradInputImages_strideWidth = gradInputImages->stride[2];\n\n int grids_strideBatch = grids->stride[0];\n int grids_strideHeight = grids->stride[1];\n int grids_strideWidth = grids->stride[2];\n\n int gradGrids_strideBatch = gradGrids->stride[0];\n int gradGrids_strideHeight = gradGrids->stride[1];\n int gradGrids_strideWidth = gradGrids->stride[2];\n\n real *inputImages_data, *gradOutput_data, *grids_data, *gradGrids_data, *gradInputImages_data;\n inputImages_data = THFloatTensor_data(inputImages);\n gradOutput_data = THFloatTensor_data(gradOutput);\n grids_data = THFloatTensor_data(grids);\n gradGrids_data = THFloatTensor_data(gradGrids);\n gradInputImages_data = THFloatTensor_data(gradInputImages);\n\n int b, yOut, xOut;\n\n for(b=0; b < batchsize; b++)\n {\n for(yOut=0; yOut < gradOutput_height; yOut++)\n {\n for(xOut=0; xOut < gradOutput_width; xOut++)\n {\n //read the grid\n real yf = grids_data[b*grids_strideBatch + yOut*grids_strideHeight + xOut*grids_strideWidth];\n real xf = grids_data[b*grids_strideBatch + yOut*grids_strideHeight + xOut*grids_strideWidth + 1];\n\n // get the weights for interpolation\n int yInTopLeft, xInTopLeft;\n real yWeightTopLeft, xWeightTopLeft;\n\n real xcoord = (xf + 1) * (inputImages_width - 1) / 2;\n xInTopLeft = floor(xcoord);\n xWeightTopLeft = 1 - (xcoord - xInTopLeft);\n\n real ycoord = (yf + 1) * (inputImages_height - 1) / 2;\n yInTopLeft = floor(ycoord);\n yWeightTopLeft = 1 - (ycoord - yInTopLeft);\n\n\n const int inTopLeftAddress = inputImages_strideBatch * b + inputImages_strideHeight * yInTopLeft + inputImages_strideWidth * xInTopLeft;\n const int inTopRightAddress = inTopLeftAddress + inputImages_strideWidth;\n const int inBottomLeftAddress = inTopLeftAddress + inputImages_strideHeight;\n const int inBottomRightAddress = inBottomLeftAddress + inputImages_strideWidth;\n\n const int gradInputImagesTopLeftAddress = gradInputImages_strideBatch * b + gradInputImages_strideHeight * yInTopLeft + gradInputImages_strideWidth * xInTopLeft;\n const int gradInputImagesTopRightAddress = gradInputImagesTopLeftAddress + gradInputImages_strideWidth;\n const int gradInputImagesBottomLeftAddress = gradInputImagesTopLeftAddress + gradInputImages_strideHeight;\n const int gradInputImagesBottomRightAddress = gradInputImagesBottomLeftAddress + gradInputImages_strideWidth;\n\n const int gradOutputAddress = gradOutput_strideBatch * b + gradOutput_strideHeight * yOut + gradOutput_strideWidth * xOut;\n\n real topLeftDotProduct = 0;\n real topRightDotProduct = 0;\n real bottomLeftDotProduct = 0;\n real bottomRightDotProduct = 0;\n\n real v=0;\n real inTopLeft=0;\n real inTopRight=0;\n real inBottomLeft=0;\n real inBottomRight=0;\n\n // we are careful with the boundaries\n bool topLeftIsIn = xInTopLeft >= 0 && xInTopLeft <= inputImages_width-1 && yInTopLeft >= 0 && yInTopLeft <= inputImages_height-1;\n bool topRightIsIn = xInTopLeft+1 >= 0 && xInTopLeft+1 <= inputImages_width-1 && yInTopLeft >= 0 && yInTopLeft <= inputImages_height-1;\n bool bottomLeftIsIn = xInTopLeft >= 0 && xInTopLeft <= inputImages_width-1 && yInTopLeft+1 >= 0 && yInTopLeft+1 <= inputImages_height-1;\n bool bottomRightIsIn = xInTopLeft+1 >= 0 && xInTopLeft+1 <= inputImages_width-1 && yInTopLeft+1 >= 0 && yInTopLeft+1 <= inputImages_height-1;\n\n int t;\n\n for(t=0; tsize[0];\n int inputImages_height = inputImages->size[2];\n int inputImages_width = inputImages->size[3];\n \n int output_height = output->size[2];\n int output_width = output->size[3];\n int inputImages_channels = inputImages->size[1];\n\n int output_strideBatch = output->stride[0];\n int output_strideHeight = output->stride[2];\n int output_strideWidth = output->stride[3]; \n int output_strideChannel = output->stride[1];\n \n\n int inputImages_strideBatch = inputImages->stride[0];\n int inputImages_strideHeight = inputImages->stride[2];\n int inputImages_strideWidth = inputImages->stride[3];\n int inputImages_strideChannel = inputImages->stride[1];\n\n int grids_strideBatch = grids->stride[0];\n int grids_strideHeight = grids->stride[2];\n int grids_strideWidth = grids->stride[3];\n int grids_strideChannel = grids->stride[1];\n\n\n real *inputImages_data, *output_data, *grids_data;\n inputImages_data = THFloatTensor_data(inputImages);\n output_data = THFloatTensor_data(output);\n grids_data = THFloatTensor_data(grids);\n\n int b, yOut, xOut;\n\n for(b=0; b < batchsize; b++)\n {\n for(yOut=0; yOut < output_height; yOut++)\n {\n for(xOut=0; xOut < output_width; xOut++)\n {\n //read the grid\n \n real xf = grids_data[b*grids_strideBatch + yOut*grids_strideHeight + xOut*grids_strideWidth + grids_strideChannel];\n real yf = grids_data[b*grids_strideBatch + yOut*grids_strideHeight + xOut*grids_strideWidth];\n\n // get the weights for interpolation\n int yInTopLeft, xInTopLeft;\n real yWeightTopLeft, xWeightTopLeft;\n\n real xcoord = (xf + 1) * (inputImages_width - 1) / 2;\n xInTopLeft = floor(xcoord);\n xWeightTopLeft = 1 - (xcoord - xInTopLeft);\n\n real ycoord = (yf + 1) * (inputImages_height - 1) / 2;\n yInTopLeft = floor(ycoord);\n yWeightTopLeft = 1 - (ycoord - yInTopLeft);\n\n\n\n const int outAddress = output_strideBatch * b + output_strideHeight * yOut + output_strideWidth * xOut;\n const int inTopLeftAddress = inputImages_strideBatch * b + inputImages_strideHeight * yInTopLeft + inputImages_strideWidth * xInTopLeft;\n const int inTopRightAddress = inTopLeftAddress + inputImages_strideWidth;\n const int inBottomLeftAddress = inTopLeftAddress + inputImages_strideHeight;\n const int inBottomRightAddress = inBottomLeftAddress + inputImages_strideWidth;\n\n real v=0;\n real inTopLeft=0;\n real inTopRight=0;\n real inBottomLeft=0;\n real inBottomRight=0;\n\n // we are careful with the boundaries\n bool topLeftIsIn = xInTopLeft >= 0 && xInTopLeft <= inputImages_width-1 && yInTopLeft >= 0 && yInTopLeft <= inputImages_height-1;\n bool topRightIsIn = xInTopLeft+1 >= 0 && xInTopLeft+1 <= inputImages_width-1 && yInTopLeft >= 0 && yInTopLeft <= inputImages_height-1;\n bool bottomLeftIsIn = xInTopLeft >= 0 && xInTopLeft <= inputImages_width-1 && yInTopLeft+1 >= 0 && yInTopLeft+1 <= inputImages_height-1;\n bool bottomRightIsIn = xInTopLeft+1 >= 0 && xInTopLeft+1 <= inputImages_width-1 && yInTopLeft+1 >= 0 && yInTopLeft+1 <= inputImages_height-1;\n\n int t;\n // interpolation happens here\n for(t=0; tsize[0];\n int inputImages_height = inputImages->size[2];\n int inputImages_width = inputImages->size[3];\n int gradOutput_height = gradOutput->size[2];\n int gradOutput_width = gradOutput->size[3];\n int inputImages_channels = inputImages->size[1];\n\n int gradOutput_strideBatch = gradOutput->stride[0];\n int gradOutput_strideHeight = gradOutput->stride[2];\n int gradOutput_strideWidth = gradOutput->stride[3];\n int gradOutput_strideChannel = gradOutput->stride[1];\n\n int inputImages_strideBatch = inputImages->stride[0];\n int inputImages_strideHeight = inputImages->stride[2];\n int inputImages_strideWidth = inputImages->stride[3];\n int inputImages_strideChannel = inputImages->stride[1];\n \n\n int gradInputImages_strideBatch = gradInputImages->stride[0];\n int gradInputImages_strideHeight = gradInputImages->stride[2];\n int gradInputImages_strideWidth = gradInputImages->stride[3];\n int gradInputImages_strideChannel = gradInputImages->stride[1];\n\n int grids_strideBatch = grids->stride[0];\n int grids_strideHeight = grids->stride[2];\n int grids_strideWidth = grids->stride[3];\n int grids_strideChannel = grids->stride[1];\n\n int gradGrids_strideBatch = gradGrids->stride[0];\n int gradGrids_strideHeight = gradGrids->stride[2];\n int gradGrids_strideWidth = gradGrids->stride[3];\n int gradGrids_strideChannel = gradGrids->stride[1];\n\n real *inputImages_data, *gradOutput_data, *grids_data, *gradGrids_data, *gradInputImages_data;\n inputImages_data = THFloatTensor_data(inputImages);\n gradOutput_data = THFloatTensor_data(gradOutput);\n grids_data = THFloatTensor_data(grids);\n gradGrids_data = THFloatTensor_data(gradGrids);\n gradInputImages_data = THFloatTensor_data(gradInputImages);\n\n int b, yOut, xOut;\n\n for(b=0; b < batchsize; b++)\n {\n for(yOut=0; yOut < gradOutput_height; yOut++)\n {\n for(xOut=0; xOut < gradOutput_width; xOut++)\n {\n //read the grid\n real xf = grids_data[b*grids_strideBatch + yOut*grids_strideHeight + xOut*grids_strideWidth + grids_strideChannel];\n real yf = grids_data[b*grids_strideBatch + yOut*grids_strideHeight + xOut*grids_strideWidth];\n \n // get the weights for interpolation\n int yInTopLeft, xInTopLeft;\n real yWeightTopLeft, xWeightTopLeft;\n\n real xcoord = (xf + 1) * (inputImages_width - 1) / 2;\n xInTopLeft = floor(xcoord);\n xWeightTopLeft = 1 - (xcoord - xInTopLeft);\n\n real ycoord = (yf + 1) * (inputImages_height - 1) / 2;\n yInTopLeft = floor(ycoord);\n yWeightTopLeft = 1 - (ycoord - yInTopLeft);\n\n\n const int inTopLeftAddress = inputImages_strideBatch * b + inputImages_strideHeight * yInTopLeft + inputImages_strideWidth * xInTopLeft;\n const int inTopRightAddress = inTopLeftAddress + inputImages_strideWidth;\n const int inBottomLeftAddress = inTopLeftAddress + inputImages_strideHeight;\n const int inBottomRightAddress = inBottomLeftAddress + inputImages_strideWidth;\n\n const int gradInputImagesTopLeftAddress = gradInputImages_strideBatch * b + gradInputImages_strideHeight * yInTopLeft + gradInputImages_strideWidth * xInTopLeft;\n const int gradInputImagesTopRightAddress = gradInputImagesTopLeftAddress + gradInputImages_strideWidth;\n const int gradInputImagesBottomLeftAddress = gradInputImagesTopLeftAddress + gradInputImages_strideHeight;\n const int gradInputImagesBottomRightAddress = gradInputImagesBottomLeftAddress + gradInputImages_strideWidth;\n\n const int gradOutputAddress = gradOutput_strideBatch * b + gradOutput_strideHeight * yOut + gradOutput_strideWidth * xOut;\n\n real topLeftDotProduct = 0;\n real topRightDotProduct = 0;\n real bottomLeftDotProduct = 0;\n real bottomRightDotProduct = 0;\n\n real v=0;\n real inTopLeft=0;\n real inTopRight=0;\n real inBottomLeft=0;\n real inBottomRight=0;\n\n // we are careful with the boundaries\n bool topLeftIsIn = xInTopLeft >= 0 && xInTopLeft <= inputImages_width-1 && yInTopLeft >= 0 && yInTopLeft <= inputImages_height-1;\n bool topRightIsIn = xInTopLeft+1 >= 0 && xInTopLeft+1 <= inputImages_width-1 && yInTopLeft >= 0 && yInTopLeft <= inputImages_height-1;\n bool bottomLeftIsIn = xInTopLeft >= 0 && xInTopLeft <= inputImages_width-1 && yInTopLeft+1 >= 0 && yInTopLeft+1 <= inputImages_height-1;\n bool bottomRightIsIn = xInTopLeft+1 >= 0 && xInTopLeft+1 <= inputImages_width-1 && yInTopLeft+1 >= 0 && yInTopLeft+1 <= inputImages_height-1;\n\n int t;\n\n for(t=0; t\n#include \n#include \n#include \"roi_crop_cuda_kernel.h\"\n\n#define real float\n\n// this symbol will be resolved automatically from PyTorch libs\nextern THCState *state;\n\n// Bilinear sampling is done in BHWD (coalescing is not obvious in BDHW)\n// we assume BHWD format in inputImages\n// we assume BHW(YX) format on grids\n\nint BilinearSamplerBHWD_updateOutput_cuda(THCudaTensor *inputImages, THCudaTensor *grids, THCudaTensor *output){\n// THCState *state = getCutorchState(L);\n// THCudaTensor *inputImages = (THCudaTensor *)luaT_checkudata(L, 2, \"torch.CudaTensor\");\n// THCudaTensor *grids = (THCudaTensor *)luaT_checkudata(L, 3, \"torch.CudaTensor\");\n// THCudaTensor *output = (THCudaTensor *)luaT_checkudata(L, 4, \"torch.CudaTensor\");\n\n int success = 0;\n success = BilinearSamplerBHWD_updateOutput_cuda_kernel(output->size[1],\n output->size[3],\n output->size[2],\n output->size[0],\n THCudaTensor_size(state, inputImages, 1),\n THCudaTensor_size(state, inputImages, 2),\n THCudaTensor_size(state, inputImages, 3),\n THCudaTensor_size(state, inputImages, 0),\n THCudaTensor_data(state, inputImages),\n THCudaTensor_stride(state, inputImages, 0),\n THCudaTensor_stride(state, inputImages, 1),\n THCudaTensor_stride(state, inputImages, 2),\n THCudaTensor_stride(state, inputImages, 3),\n THCudaTensor_data(state, grids),\n THCudaTensor_stride(state, grids, 0),\n THCudaTensor_stride(state, grids, 3),\n THCudaTensor_stride(state, grids, 1),\n THCudaTensor_stride(state, grids, 2),\n THCudaTensor_data(state, output),\n THCudaTensor_stride(state, output, 0),\n THCudaTensor_stride(state, output, 1),\n THCudaTensor_stride(state, output, 2),\n THCudaTensor_stride(state, output, 3),\n THCState_getCurrentStream(state));\n\n //check for errors\n if (!success) {\n THError(\"aborting\");\n }\n return 1;\n}\n\nint BilinearSamplerBHWD_updateGradInput_cuda(THCudaTensor *inputImages, THCudaTensor *grids, THCudaTensor *gradInputImages,\n THCudaTensor *gradGrids, THCudaTensor *gradOutput)\n{\n// THCState *state = getCutorchState(L);\n// THCudaTensor *inputImages = (THCudaTensor *)luaT_checkudata(L, 2, \"torch.CudaTensor\");\n// THCudaTensor *grids = (THCudaTensor *)luaT_checkudata(L, 3, \"torch.CudaTensor\");\n// THCudaTensor *gradInputImages = (THCudaTensor *)luaT_checkudata(L, 4, \"torch.CudaTensor\");\n// THCudaTensor *gradGrids = (THCudaTensor *)luaT_checkudata(L, 5, \"torch.CudaTensor\");\n// THCudaTensor *gradOutput = (THCudaTensor *)luaT_checkudata(L, 6, \"torch.CudaTensor\");\n\n int success = 0;\n success = BilinearSamplerBHWD_updateGradInput_cuda_kernel(gradOutput->size[1],\n gradOutput->size[3],\n gradOutput->size[2],\n gradOutput->size[0],\n THCudaTensor_size(state, inputImages, 1),\n THCudaTensor_size(state, inputImages, 2),\n THCudaTensor_size(state, inputImages, 3),\n THCudaTensor_size(state, inputImages, 0),\n THCudaTensor_data(state, inputImages),\n THCudaTensor_stride(state, inputImages, 0),\n THCudaTensor_stride(state, inputImages, 1),\n THCudaTensor_stride(state, inputImages, 2),\n THCudaTensor_stride(state, inputImages, 3),\n THCudaTensor_data(state, grids),\n THCudaTensor_stride(state, grids, 0),\n THCudaTensor_stride(state, grids, 3),\n THCudaTensor_stride(state, grids, 1),\n THCudaTensor_stride(state, grids, 2),\n THCudaTensor_data(state, gradInputImages),\n THCudaTensor_stride(state, gradInputImages, 0),\n THCudaTensor_stride(state, gradInputImages, 1),\n THCudaTensor_stride(state, gradInputImages, 2),\n THCudaTensor_stride(state, gradInputImages, 3),\n THCudaTensor_data(state, gradGrids),\n THCudaTensor_stride(state, gradGrids, 0),\n THCudaTensor_stride(state, gradGrids, 3),\n THCudaTensor_stride(state, gradGrids, 1),\n THCudaTensor_stride(state, gradGrids, 2),\n THCudaTensor_data(state, gradOutput),\n THCudaTensor_stride(state, gradOutput, 0),\n THCudaTensor_stride(state, gradOutput, 1),\n THCudaTensor_stride(state, gradOutput, 2),\n THCudaTensor_stride(state, gradOutput, 3),\n THCState_getCurrentStream(state));\n\n //check for errors\n if (!success) {\n THError(\"aborting\");\n }\n return 1;\n}\n"
},
{
"path": "lib/model/roi_crop/src/roi_crop_cuda.h",
"content": "// Bilinear sampling is done in BHWD (coalescing is not obvious in BDHW)\n// we assume BHWD format in inputImages\n// we assume BHW(YX) format on grids\n\nint BilinearSamplerBHWD_updateOutput_cuda(THCudaTensor *inputImages, THCudaTensor *grids, THCudaTensor *output);\n\nint BilinearSamplerBHWD_updateGradInput_cuda(THCudaTensor *inputImages, THCudaTensor *grids, THCudaTensor *gradInputImages,\n THCudaTensor *gradGrids, THCudaTensor *gradOutput);\n"
},
{
"path": "lib/model/roi_crop/src/roi_crop_cuda_kernel.cu",
"content": "#include \n#include \n#include \"roi_crop_cuda_kernel.h\"\n\n#define real float\n\n// Bilinear sampling is done in BHWD (coalescing is not obvious in BDHW)\n// we assume BHWD format in inputImages\n// we assume BHW(YX) format on grids\n\n__device__ void getTopLeft(float x, int width, int& point, float& weight)\n{\n /* for interpolation :\n stores in point and weight :\n - the x-coordinate of the pixel on the left (or y-coordinate of the upper pixel)\n - the weight for interpolating\n */\n\n float xcoord = (x + 1) * (width - 1) / 2;\n point = floor(xcoord);\n weight = 1 - (xcoord - point);\n}\n\n__device__ bool between(int value, int lowerBound, int upperBound)\n{\n return (value >= lowerBound && value <= upperBound);\n}\n\n__device__ void sumReduceShMem(volatile float s[])\n{\n /* obviously only works for 32 elements */\n /* sums up a shared memory array of 32 elements, stores it in s[0] */\n /* whole warp can then read first element (broadcasting) */\n if(threadIdx.x<16) { s[threadIdx.x] = s[threadIdx.x] + s[threadIdx.x+16]; }\n if(threadIdx.x<8) { s[threadIdx.x] = s[threadIdx.x] + s[threadIdx.x+8]; }\n if(threadIdx.x<4) { s[threadIdx.x] = s[threadIdx.x] + s[threadIdx.x+4]; }\n if(threadIdx.x<2) { s[threadIdx.x] = s[threadIdx.x] + s[threadIdx.x+2]; }\n if(threadIdx.x<1) { s[threadIdx.x] = s[threadIdx.x] + s[threadIdx.x+1]; }\n}\n\n// CUDA: grid stride looping\n#define CUDA_KERNEL_LOOP(i, n) \\\n for (int i = blockIdx.x * blockDim.x + threadIdx.x; \\\n i < (n); \\\n i += blockDim.x * gridDim.x)\n\n__global__ void bilinearSamplingFromGrid(const int nthreads, float* inputImages_data, int inputImages_strideBatch, int inputImages_strideChannels, int inputImages_strideHeight, int inputImages_strideWidth,\n float* grids_data, int grids_strideBatch, int grids_strideYX, int grids_strideHeight, int grids_strideWidth,\n float* output_data, int output_strideBatch, int output_strideChannels, int output_strideHeight, int output_strideWidth,\n int inputImages_channels, int inputImages_height, int inputImages_width,\n int output_channels, int output_height, int output_width, int output_batchsize,\n int roiPerImage)\n{\n CUDA_KERNEL_LOOP(index, nthreads)\n {\n const int xOut = index % output_width;\n const int yOut = (index / output_width) % output_height;\n const int cOut = (index / output_width / output_height) % output_channels;\n const int b = index / output_width / output_height / output_channels;\n\n const int width = inputImages_width;\n const int height = inputImages_height;\n\n const int b_input = b / roiPerImage;\n\n float yf = grids_data[b*grids_strideBatch + yOut*grids_strideHeight + xOut*grids_strideWidth];\n float xf = grids_data[b*grids_strideBatch + yOut*grids_strideHeight + xOut*grids_strideWidth + 1];\n\n int yInTopLeft, xInTopLeft;\n float yWeightTopLeft, xWeightTopLeft;\n getTopLeft(xf, inputImages_width, xInTopLeft, xWeightTopLeft);\n getTopLeft(yf, inputImages_height, yInTopLeft, yWeightTopLeft);\n\n // const int outAddress = output_strideBatch * b + output_strideHeight * yOut + output_strideWidth * xOut;\n const int outAddress = output_strideBatch * b + output_strideChannels * cOut + output_strideHeight * yOut + xOut;\n\n const int inTopLeftAddress = inputImages_strideBatch * b_input + inputImages_strideChannels * cOut + inputImages_strideHeight * yInTopLeft + xInTopLeft;\n const int inTopRightAddress = inTopLeftAddress + inputImages_strideWidth;\n const int inBottomLeftAddress = inTopLeftAddress + inputImages_strideHeight;\n const int inBottomRightAddress = inBottomLeftAddress + inputImages_strideWidth;\n\n float v=0;\n float inTopLeft=0;\n float inTopRight=0;\n float inBottomLeft=0;\n float inBottomRight=0;\n\n bool topLeftIsIn = between(xInTopLeft, 0, width-1) && between(yInTopLeft, 0, height-1);\n bool topRightIsIn = between(xInTopLeft+1, 0, width-1) && between(yInTopLeft, 0, height-1);\n bool bottomLeftIsIn = between(xInTopLeft, 0, width-1) && between(yInTopLeft+1, 0, height-1);\n bool bottomRightIsIn = between(xInTopLeft+1, 0, width-1) && between(yInTopLeft+1, 0, height-1);\n\n if (!topLeftIsIn && !topRightIsIn && !bottomLeftIsIn && !bottomRightIsIn)\n continue;\n\n if(topLeftIsIn) inTopLeft = inputImages_data[inTopLeftAddress];\n if(topRightIsIn) inTopRight = inputImages_data[inTopRightAddress];\n if(bottomLeftIsIn) inBottomLeft = inputImages_data[inBottomLeftAddress];\n if(bottomRightIsIn) inBottomRight = inputImages_data[inBottomRightAddress];\n\n v = xWeightTopLeft * yWeightTopLeft * inTopLeft\n + (1 - xWeightTopLeft) * yWeightTopLeft * inTopRight\n + xWeightTopLeft * (1 - yWeightTopLeft) * inBottomLeft\n + (1 - xWeightTopLeft) * (1 - yWeightTopLeft) * inBottomRight;\n\n output_data[outAddress] = v;\n }\n\n}\n\n__global__ void backwardBilinearSampling(const int nthreads, float* inputImages_data, int inputImages_strideBatch, int inputImages_strideChannels, int inputImages_strideHeight, int inputImages_strideWidth,\n float* gradInputImages_data, int gradInputImages_strideBatch, int gradInputImages_strideChannels, int gradInputImages_strideHeight, int gradInputImages_strideWidth,\n float* grids_data, int grids_strideBatch, int grids_strideYX, int grids_strideHeight, int grids_strideWidth,\n float* gradGrids_data, int gradGrids_strideBatch, int gradGrids_strideYX, int gradGrids_strideHeight, int gradGrids_strideWidth,\n float* gradOutput_data, int gradOutput_strideBatch, int gradOutput_strideChannels, int gradOutput_strideHeight, int gradOutput_strideWidth,\n int inputImages_channels, int inputImages_height, int inputImages_width,\n int gradOutput_channels, int gradOutput_height, int gradOutput_width, int gradOutput_batchsize,\n int roiPerImage)\n{\n\n CUDA_KERNEL_LOOP(index, nthreads)\n {\n const int xOut = index % gradOutput_width;\n const int yOut = (index / gradOutput_width) % gradOutput_height;\n const int cOut = (index / gradOutput_width / gradOutput_height) % gradOutput_channels;\n const int b = index / gradOutput_width / gradOutput_height / gradOutput_channels;\n\n const int b_input = b / roiPerImage;\n\n const int width = inputImages_width;\n const int height = inputImages_height;\n\n float yf = grids_data[b*grids_strideBatch + yOut*grids_strideHeight + xOut*grids_strideWidth];\n float xf = grids_data[b*grids_strideBatch + yOut*grids_strideHeight + xOut*grids_strideWidth + 1];\n\n int yInTopLeft, xInTopLeft;\n float yWeightTopLeft, xWeightTopLeft;\n getTopLeft(xf, inputImages_width, xInTopLeft, xWeightTopLeft);\n getTopLeft(yf, inputImages_height, yInTopLeft, yWeightTopLeft);\n\n const int inTopLeftAddress = inputImages_strideBatch * b_input + inputImages_strideChannels * cOut + inputImages_strideHeight * yInTopLeft + xInTopLeft;\n const int inTopRightAddress = inTopLeftAddress + inputImages_strideWidth;\n const int inBottomLeftAddress = inTopLeftAddress + inputImages_strideHeight;\n const int inBottomRightAddress = inBottomLeftAddress + inputImages_strideWidth;\n\n const int gradInputImagesTopLeftAddress = gradInputImages_strideBatch * b_input + gradInputImages_strideChannels * cOut\n + gradInputImages_strideHeight * yInTopLeft + xInTopLeft;\n const int gradInputImagesTopRightAddress = gradInputImagesTopLeftAddress + gradInputImages_strideWidth;\n const int gradInputImagesBottomLeftAddress = gradInputImagesTopLeftAddress + gradInputImages_strideHeight;\n const int gradInputImagesBottomRightAddress = gradInputImagesBottomLeftAddress + gradInputImages_strideWidth;\n\n const int gradOutputAddress = gradOutput_strideBatch * b + gradOutput_strideChannels * cOut + gradOutput_strideHeight * yOut + xOut;\n\n float topLeftDotProduct = 0;\n float topRightDotProduct = 0;\n float bottomLeftDotProduct = 0;\n float bottomRightDotProduct = 0;\n\n bool topLeftIsIn = between(xInTopLeft, 0, width-1) && between(yInTopLeft, 0, height-1);\n bool topRightIsIn = between(xInTopLeft+1, 0, width-1) && between(yInTopLeft, 0, height-1);\n bool bottomLeftIsIn = between(xInTopLeft, 0, width-1) && between(yInTopLeft+1, 0, height-1);\n bool bottomRightIsIn = between(xInTopLeft+1, 0, width-1) && between(yInTopLeft+1, 0, height-1);\n\n float gradOutValue = gradOutput_data[gradOutputAddress];\n // bool between(int value, int lowerBound, int upperBound)\n if(topLeftIsIn)\n {\n float inTopLeft = inputImages_data[inTopLeftAddress];\n topLeftDotProduct += inTopLeft * gradOutValue;\n atomicAdd(&gradInputImages_data[gradInputImagesTopLeftAddress], xWeightTopLeft * yWeightTopLeft * gradOutValue);\n }\n\n if(topRightIsIn)\n {\n float inTopRight = inputImages_data[inTopRightAddress];\n topRightDotProduct += inTopRight * gradOutValue;\n atomicAdd(&gradInputImages_data[gradInputImagesTopRightAddress], (1 - xWeightTopLeft) * yWeightTopLeft * gradOutValue);\n }\n\n if(bottomLeftIsIn)\n {\n float inBottomLeft = inputImages_data[inBottomLeftAddress];\n bottomLeftDotProduct += inBottomLeft * gradOutValue;\n atomicAdd(&gradInputImages_data[gradInputImagesBottomLeftAddress], xWeightTopLeft * (1 - yWeightTopLeft) * gradOutValue);\n }\n\n if(bottomRightIsIn)\n {\n float inBottomRight = inputImages_data[inBottomRightAddress];\n bottomRightDotProduct += inBottomRight * gradOutValue;\n atomicAdd(&gradInputImages_data[gradInputImagesBottomRightAddress], (1 - xWeightTopLeft) * (1 - yWeightTopLeft) * gradOutValue);\n }\n }\n}\n\n\n#ifdef __cplusplus\nextern \"C\" {\n#endif\n\nint BilinearSamplerBHWD_updateOutput_cuda_kernel(/*output->size[1]*/int oc,\n /*output->size[3]*/int ow,\n /*output->size[2]*/int oh,\n /*output->size[0]*/int ob,\n /*THCudaTensor_size(state, inputImages, 1)*/int ic,\n /*THCudaTensor_size(state, inputImages, 2)*/int ih,\n /*THCudaTensor_size(state, inputImages, 3)*/int iw,\n /*THCudaTensor_size(state, inputImages, 0)*/int ib,\n /*THCudaTensor *inputImages*/float *inputImages, int isb, int isc, int ish, int isw,\n /*THCudaTensor *grids*/float *grids, int gsb, int gsc, int gsh, int gsw,\n /*THCudaTensor *output*/float *output, int osb, int osc, int osh, int osw,\n /*THCState_getCurrentStream(state)*/cudaStream_t stream)\n{\n const int kThreadsPerBlock = 1024;\n int output_size = ob * oh * ow * oc;\n cudaError_t err;\n int roiPerImage = ob / ib;\n\n // printf(\"forward pass\\n\");\n\n bilinearSamplingFromGrid<<<(output_size + kThreadsPerBlock - 1) / kThreadsPerBlock, kThreadsPerBlock, 0, stream>>>(\n output_size,\n /*THCudaTensor_data(state, inputImages)*/inputImages,\n /*THCudaTensor_stride(state, inputImages, 0)*/isb,\n /*THCudaTensor_stride(state, inputImages, 3)*/isc,\n /*THCudaTensor_stride(state, inputImages, 1)*/ish,\n /*THCudaTensor_stride(state, inputImages, 2)*/isw,\n /*THCudaTensor_data(state, grids)*/grids,\n /*THCudaTensor_stride(state, grids, 0)*/gsb,\n /*THCudaTensor_stride(state, grids, 3)*/gsc,\n /*THCudaTensor_stride(state, grids, 1)*/gsh,\n /*THCudaTensor_stride(state, grids, 2)*/gsw,\n /*THCudaTensor_data(state, output)*/output,\n /*THCudaTensor_stride(state, output, 0)*/osb,\n /*THCudaTensor_stride(state, output, 3)*/osc,\n /*THCudaTensor_stride(state, output, 1)*/osh,\n /*THCudaTensor_stride(state, output, 2)*/osw,\n /*THCudaTensor_size(state, inputImages, 3)*/ic,\n /*THCudaTensor_size(state, inputImages, 1)*/ih,\n /*THCudaTensor_size(state, inputImages, 2)*/iw,\n /*THCudaTensor_size(state, output, 3)*/oc,\n /*THCudaTensor_size(state, output, 1)*/oh,\n /*THCudaTensor_size(state, output, 2)*/ow,\n /*THCudaTensor_size(state, output, 0)*/ob,\n /*Number of rois per image*/roiPerImage);\n\n // check for errors\n err = cudaGetLastError();\n if (err != cudaSuccess) {\n printf(\"error in BilinearSampler.updateOutput: %s\\n\", cudaGetErrorString(err));\n //THError(\"aborting\");\n return 0;\n }\n return 1;\n}\n\nint BilinearSamplerBHWD_updateGradInput_cuda_kernel(/*gradOutput->size[1]*/int goc,\n /*gradOutput->size[3]*/int gow,\n /*gradOutput->size[2]*/int goh,\n /*gradOutput->size[0]*/int gob,\n /*THCudaTensor_size(state, inputImages, 1)*/int ic,\n /*THCudaTensor_size(state, inputImages, 2)*/int ih,\n /*THCudaTensor_size(state, inputImages, 3)*/int iw,\n /*THCudaTensor_size(state, inputImages, 0)*/int ib,\n /*THCudaTensor *inputImages*/float *inputImages, int isb, int isc, int ish, int isw,\n /*THCudaTensor *grids*/float *grids, int gsb, int gsc, int gsh, int gsw,\n /*THCudaTensor *gradInputImages*/float *gradInputImages, int gisb, int gisc, int gish, int gisw,\n /*THCudaTensor *gradGrids*/float *gradGrids, int ggsb, int ggsc, int ggsh, int ggsw,\n /*THCudaTensor *gradOutput*/float *gradOutput, int gosb, int gosc, int gosh, int gosw,\n /*THCState_getCurrentStream(state)*/cudaStream_t stream)\n{\n\n const int kThreadsPerBlock = 1024;\n int output_size = gob * goh * gow * goc;\n cudaError_t err;\n int roiPerImage = gob / ib;\n\n // printf(\"%d %d %d %d\\n\", gob, goh, gow, goc);\n // printf(\"%d %d %d %d\\n\", ib, ih, iw, ic);\n // printf(\"backward pass\\n\");\n\n backwardBilinearSampling<<<(output_size + kThreadsPerBlock - 1) / kThreadsPerBlock, kThreadsPerBlock, 0, stream>>>(\n output_size,\n /*THCudaTensor_data(state, inputImages)*/inputImages,\n /*THCudaTensor_stride(state, inputImages, 0)*/isb,\n /*THCudaTensor_stride(state, inputImages, 3)*/isc,\n /*THCudaTensor_stride(state, inputImages, 1)*/ish,\n /*THCudaTensor_stride(state, inputImages, 2)*/isw,\n /*THCudaTensor_data(state, gradInputImages)*/gradInputImages,\n /*THCudaTensor_stride(state, gradInputImages, 0)*/gisb,\n /*THCudaTensor_stride(state, gradInputImages, 3)*/gisc,\n /*THCudaTensor_stride(state, gradInputImages, 1)*/gish,\n /*THCudaTensor_stride(state, gradInputImages, 2)*/gisw,\n /*THCudaTensor_data(state, grids)*/grids,\n /*THCudaTensor_stride(state, grids, 0)*/gsb,\n /*THCudaTensor_stride(state, grids, 3)*/gsc,\n /*THCudaTensor_stride(state, grids, 1)*/gsh,\n /*THCudaTensor_stride(state, grids, 2)*/gsw,\n /*THCudaTensor_data(state, gradGrids)*/gradGrids,\n /*THCudaTensor_stride(state, gradGrids, 0)*/ggsb,\n /*THCudaTensor_stride(state, gradGrids, 3)*/ggsc,\n /*THCudaTensor_stride(state, gradGrids, 1)*/ggsh,\n /*THCudaTensor_stride(state, gradGrids, 2)*/ggsw,\n /*THCudaTensor_data(state, gradOutput)*/gradOutput,\n /*THCudaTensor_stride(state, gradOutput, 0)*/gosb,\n /*THCudaTensor_stride(state, gradOutput, 3)*/gosc,\n /*THCudaTensor_stride(state, gradOutput, 1)*/gosh,\n /*THCudaTensor_stride(state, gradOutput, 2)*/gosw,\n /*THCudaTensor_size(state, inputImages, 3)*/ic,\n /*THCudaTensor_size(state, inputImages, 1)*/ih,\n /*THCudaTensor_size(state, inputImages, 2)*/iw,\n /*THCudaTensor_size(state, gradOutput, 3)*/goc,\n /*THCudaTensor_size(state, gradOutput, 1)*/goh,\n /*THCudaTensor_size(state, gradOutput, 2)*/gow,\n /*THCudaTensor_size(state, gradOutput, 0)*/gob,\n /*Number of rois per image*/roiPerImage);\n\n // check for errors\n err = cudaGetLastError();\n if (err != cudaSuccess) {\n printf(\"error in BilinearSampler.updateGradInput: %s\\n\", cudaGetErrorString(err));\n //THError(\"aborting\");\n return 0;\n }\n return 1;\n}\n\n#ifdef __cplusplus\n}\n#endif\n"
},
{
"path": "lib/model/roi_crop/src/roi_crop_cuda_kernel.h",
"content": "#ifdef __cplusplus\nextern \"C\" {\n#endif\n\n\nint BilinearSamplerBHWD_updateOutput_cuda_kernel(/*output->size[3]*/int oc,\n /*output->size[2]*/int ow,\n /*output->size[1]*/int oh,\n /*output->size[0]*/int ob,\n /*THCudaTensor_size(state, inputImages, 3)*/int ic,\n /*THCudaTensor_size(state, inputImages, 1)*/int ih,\n /*THCudaTensor_size(state, inputImages, 2)*/int iw,\n /*THCudaTensor_size(state, inputImages, 0)*/int ib,\n /*THCudaTensor *inputImages*/float *inputImages, int isb, int isc, int ish, int isw,\n /*THCudaTensor *grids*/float *grids, int gsb, int gsc, int gsh, int gsw,\n /*THCudaTensor *output*/float *output, int osb, int osc, int osh, int osw,\n /*THCState_getCurrentStream(state)*/cudaStream_t stream);\n\nint BilinearSamplerBHWD_updateGradInput_cuda_kernel(/*gradOutput->size[3]*/int goc,\n /*gradOutput->size[2]*/int gow,\n /*gradOutput->size[1]*/int goh,\n /*gradOutput->size[0]*/int gob,\n /*THCudaTensor_size(state, inputImages, 3)*/int ic,\n /*THCudaTensor_size(state, inputImages, 1)*/int ih,\n /*THCudaTensor_size(state, inputImages, 2)*/int iw,\n /*THCudaTensor_size(state, inputImages, 0)*/int ib,\n /*THCudaTensor *inputImages*/float *inputImages, int isb, int isc, int ish, int isw,\n /*THCudaTensor *grids*/float *grids, int gsb, int gsc, int gsh, int gsw,\n /*THCudaTensor *gradInputImages*/float *gradInputImages, int gisb, int gisc, int gish, int gisw,\n /*THCudaTensor *gradGrids*/float *gradGrids, int ggsb, int ggsc, int ggsh, int ggsw,\n /*THCudaTensor *gradOutput*/float *gradOutput, int gosb, int gosc, int gosh, int gosw,\n /*THCState_getCurrentStream(state)*/cudaStream_t stream);\n\n\n#ifdef __cplusplus\n}\n#endif\n"
},
{
"path": "lib/model/roi_pooling/__init__.py",
"content": ""
},
{
"path": "lib/model/roi_pooling/_ext/__init__.py",
"content": ""
},
{
"path": "lib/model/roi_pooling/_ext/roi_pooling/__init__.py",
"content": "\nfrom torch.utils.ffi import _wrap_function\nfrom ._roi_pooling import lib as _lib, ffi as _ffi\n\n__all__ = []\ndef _import_symbols(locals):\n for symbol in dir(_lib):\n fn = getattr(_lib, symbol)\n if callable(fn):\n locals[symbol] = _wrap_function(fn, _ffi)\n else:\n locals[symbol] = fn\n __all__.append(symbol)\n\n_import_symbols(locals())\n"
},
{
"path": "lib/model/roi_pooling/build.py",
"content": "from __future__ import print_function\nimport os\nimport torch\nfrom torch.utils.ffi import create_extension\n\n\nsources = ['src/roi_pooling.c']\nheaders = ['src/roi_pooling.h']\ndefines = []\nwith_cuda = False\n\nif torch.cuda.is_available():\n print('Including CUDA code.')\n sources += ['src/roi_pooling_cuda.c']\n headers += ['src/roi_pooling_cuda.h']\n defines += [('WITH_CUDA', None)]\n with_cuda = True\n\nthis_file = os.path.dirname(os.path.realpath(__file__))\nprint(this_file)\nextra_objects = ['src/roi_pooling.cu.o']\nextra_objects = [os.path.join(this_file, fname) for fname in extra_objects]\n\nffi = create_extension(\n '_ext.roi_pooling',\n headers=headers,\n sources=sources,\n define_macros=defines,\n relative_to=__file__,\n with_cuda=with_cuda,\n extra_objects=extra_objects\n)\n\nif __name__ == '__main__':\n ffi.build()\n"
},
{
"path": "lib/model/roi_pooling/functions/__init__.py",
"content": ""
},
{
"path": "lib/model/roi_pooling/functions/roi_pool.py",
"content": "import torch\nfrom torch.autograd import Function\nfrom .._ext import roi_pooling\nimport pdb\n\nclass RoIPoolFunction(Function):\n def __init__(ctx, pooled_height, pooled_width, spatial_scale):\n ctx.pooled_width = pooled_width\n ctx.pooled_height = pooled_height\n ctx.spatial_scale = spatial_scale\n ctx.feature_size = None\n\n def forward(ctx, features, rois): \n ctx.feature_size = features.size() \n batch_size, num_channels, data_height, data_width = ctx.feature_size\n num_rois = rois.size(0)\n output = features.new(num_rois, num_channels, ctx.pooled_height, ctx.pooled_width).zero_()\n ctx.argmax = features.new(num_rois, num_channels, ctx.pooled_height, ctx.pooled_width).zero_().int()\n ctx.rois = rois\n if not features.is_cuda:\n _features = features.permute(0, 2, 3, 1)\n roi_pooling.roi_pooling_forward(ctx.pooled_height, ctx.pooled_width, ctx.spatial_scale,\n _features, rois, output)\n else:\n roi_pooling.roi_pooling_forward_cuda(ctx.pooled_height, ctx.pooled_width, ctx.spatial_scale,\n features, rois, output, ctx.argmax)\n\n return output\n\n def backward(ctx, grad_output):\n assert(ctx.feature_size is not None and grad_output.is_cuda)\n batch_size, num_channels, data_height, data_width = ctx.feature_size\n grad_input = grad_output.new(batch_size, num_channels, data_height, data_width).zero_()\n\n roi_pooling.roi_pooling_backward_cuda(ctx.pooled_height, ctx.pooled_width, ctx.spatial_scale,\n grad_output, ctx.rois, grad_input, ctx.argmax)\n\n return grad_input, None\n"
},
{
"path": "lib/model/roi_pooling/modules/__init__.py",
"content": ""
},
{
"path": "lib/model/roi_pooling/modules/roi_pool.py",
"content": "from torch.nn.modules.module import Module\nfrom ..functions.roi_pool import RoIPoolFunction\n\n\nclass _RoIPooling(Module):\n def __init__(self, pooled_height, pooled_width, spatial_scale):\n super(_RoIPooling, self).__init__()\n\n self.pooled_width = int(pooled_width)\n self.pooled_height = int(pooled_height)\n self.spatial_scale = float(spatial_scale)\n\n def forward(self, features, rois):\n return RoIPoolFunction(self.pooled_height, self.pooled_width, self.spatial_scale)(features, rois)\n"
},
{
"path": "lib/model/roi_pooling/src/roi_pooling.c",
"content": "#include | \n#include \n\nint roi_pooling_forward(int pooled_height, int pooled_width, float spatial_scale,\n THFloatTensor * features, THFloatTensor * rois, THFloatTensor * output)\n{\n // Grab the input tensor\n float * data_flat = THFloatTensor_data(features);\n float * rois_flat = THFloatTensor_data(rois);\n\n float * output_flat = THFloatTensor_data(output);\n\n // Number of ROIs\n int num_rois = THFloatTensor_size(rois, 0);\n int size_rois = THFloatTensor_size(rois, 1);\n // batch size\n int batch_size = THFloatTensor_size(features, 0);\n if(batch_size != 1)\n {\n return 0;\n }\n // data height\n int data_height = THFloatTensor_size(features, 1);\n // data width\n int data_width = THFloatTensor_size(features, 2);\n // Number of channels\n int num_channels = THFloatTensor_size(features, 3);\n\n // Set all element of the output tensor to -inf.\n THFloatStorage_fill(THFloatTensor_storage(output), -1);\n\n // For each ROI R = [batch_index x1 y1 x2 y2]: max pool over R\n int index_roi = 0;\n int index_output = 0;\n int n;\n for (n = 0; n < num_rois; ++n)\n {\n int roi_batch_ind = rois_flat[index_roi + 0];\n int roi_start_w = round(rois_flat[index_roi + 1] * spatial_scale);\n int roi_start_h = round(rois_flat[index_roi + 2] * spatial_scale);\n int roi_end_w = round(rois_flat[index_roi + 3] * spatial_scale);\n int roi_end_h = round(rois_flat[index_roi + 4] * spatial_scale);\n // CHECK_GE(roi_batch_ind, 0);\n // CHECK_LT(roi_batch_ind, batch_size);\n\n int roi_height = fmaxf(roi_end_h - roi_start_h + 1, 1);\n int roi_width = fmaxf(roi_end_w - roi_start_w + 1, 1);\n float bin_size_h = (float)(roi_height) / (float)(pooled_height);\n float bin_size_w = (float)(roi_width) / (float)(pooled_width);\n\n int index_data = roi_batch_ind * data_height * data_width * num_channels;\n const int output_area = pooled_width * pooled_height;\n\n int c, ph, pw;\n for (ph = 0; ph < pooled_height; ++ph)\n {\n for (pw = 0; pw < pooled_width; ++pw)\n {\n int hstart = (floor((float)(ph) * bin_size_h));\n int wstart = (floor((float)(pw) * bin_size_w));\n int hend = (ceil((float)(ph + 1) * bin_size_h));\n int wend = (ceil((float)(pw + 1) * bin_size_w));\n\n hstart = fminf(fmaxf(hstart + roi_start_h, 0), data_height);\n hend = fminf(fmaxf(hend + roi_start_h, 0), data_height);\n wstart = fminf(fmaxf(wstart + roi_start_w, 0), data_width);\n wend = fminf(fmaxf(wend + roi_start_w, 0), data_width);\n\n const int pool_index = index_output + (ph * pooled_width + pw);\n int is_empty = (hend <= hstart) || (wend <= wstart);\n if (is_empty)\n {\n for (c = 0; c < num_channels * output_area; c += output_area)\n {\n output_flat[pool_index + c] = 0;\n }\n }\n else\n {\n int h, w, c;\n for (h = hstart; h < hend; ++h)\n {\n for (w = wstart; w < wend; ++w)\n {\n for (c = 0; c < num_channels; ++c)\n {\n const int index = (h * data_width + w) * num_channels + c;\n if (data_flat[index_data + index] > output_flat[pool_index + c * output_area])\n {\n output_flat[pool_index + c * output_area] = data_flat[index_data + index];\n }\n }\n }\n }\n }\n }\n }\n\n // Increment ROI index\n index_roi += size_rois;\n index_output += pooled_height * pooled_width * num_channels;\n }\n return 1;\n}"
},
{
"path": "lib/model/roi_pooling/src/roi_pooling.h",
"content": "int roi_pooling_forward(int pooled_height, int pooled_width, float spatial_scale,\n THFloatTensor * features, THFloatTensor * rois, THFloatTensor * output);"
},
{
"path": "lib/model/roi_pooling/src/roi_pooling_cuda.c",
"content": "#include \n#include \n#include \"roi_pooling_kernel.h\"\n\nextern THCState *state;\n\nint roi_pooling_forward_cuda(int pooled_height, int pooled_width, float spatial_scale,\n THCudaTensor * features, THCudaTensor * rois, THCudaTensor * output, THCudaIntTensor * argmax)\n{\n // Grab the input tensor\n float * data_flat = THCudaTensor_data(state, features);\n float * rois_flat = THCudaTensor_data(state, rois);\n\n float * output_flat = THCudaTensor_data(state, output);\n int * argmax_flat = THCudaIntTensor_data(state, argmax);\n\n // Number of ROIs\n int num_rois = THCudaTensor_size(state, rois, 0);\n int size_rois = THCudaTensor_size(state, rois, 1);\n if (size_rois != 5)\n {\n return 0;\n }\n\n // batch size\n // int batch_size = THCudaTensor_size(state, features, 0);\n // if (batch_size != 1)\n // {\n // return 0;\n // }\n // data height\n int data_height = THCudaTensor_size(state, features, 2);\n // data width\n int data_width = THCudaTensor_size(state, features, 3);\n // Number of channels\n int num_channels = THCudaTensor_size(state, features, 1);\n\n cudaStream_t stream = THCState_getCurrentStream(state);\n\n ROIPoolForwardLaucher(\n data_flat, spatial_scale, num_rois, data_height,\n data_width, num_channels, pooled_height,\n pooled_width, rois_flat,\n output_flat, argmax_flat, stream);\n\n return 1;\n}\n\nint roi_pooling_backward_cuda(int pooled_height, int pooled_width, float spatial_scale,\n THCudaTensor * top_grad, THCudaTensor * rois, THCudaTensor * bottom_grad, THCudaIntTensor * argmax)\n{\n // Grab the input tensor\n float * top_grad_flat = THCudaTensor_data(state, top_grad);\n float * rois_flat = THCudaTensor_data(state, rois);\n\n float * bottom_grad_flat = THCudaTensor_data(state, bottom_grad);\n int * argmax_flat = THCudaIntTensor_data(state, argmax);\n\n // Number of ROIs\n int num_rois = THCudaTensor_size(state, rois, 0);\n int size_rois = THCudaTensor_size(state, rois, 1);\n if (size_rois != 5)\n {\n return 0;\n }\n\n // batch size\n int batch_size = THCudaTensor_size(state, bottom_grad, 0);\n // if (batch_size != 1)\n // {\n // return 0;\n // }\n // data height\n int data_height = THCudaTensor_size(state, bottom_grad, 2);\n // data width\n int data_width = THCudaTensor_size(state, bottom_grad, 3);\n // Number of channels\n int num_channels = THCudaTensor_size(state, bottom_grad, 1);\n\n cudaStream_t stream = THCState_getCurrentStream(state);\n ROIPoolBackwardLaucher(\n top_grad_flat, spatial_scale, batch_size, num_rois, data_height,\n data_width, num_channels, pooled_height,\n pooled_width, rois_flat,\n bottom_grad_flat, argmax_flat, stream);\n\n return 1;\n}\n"
},
{
"path": "lib/model/roi_pooling/src/roi_pooling_cuda.h",
"content": "int roi_pooling_forward_cuda(int pooled_height, int pooled_width, float spatial_scale,\n THCudaTensor * features, THCudaTensor * rois, THCudaTensor * output, THCudaIntTensor * argmax);\n\nint roi_pooling_backward_cuda(int pooled_height, int pooled_width, float spatial_scale,\n THCudaTensor * top_grad, THCudaTensor * rois, THCudaTensor * bottom_grad, THCudaIntTensor * argmax);"
},
{
"path": "lib/model/roi_pooling/src/roi_pooling_kernel.cu",
"content": "// #ifdef __cplusplus\n// extern \"C\" {\n// #endif\n\n#include \n#include \n#include \n#include \n#include \"roi_pooling_kernel.h\"\n\n\n#define DIVUP(m, n) ((m) / (m) + ((m) % (n) > 0))\n\n#define CUDA_1D_KERNEL_LOOP(i, n) \\\n for (int i = blockIdx.x * blockDim.x + threadIdx.x; i < n; \\\n i += blockDim.x * gridDim.x)\n\n// CUDA: grid stride looping\n#define CUDA_KERNEL_LOOP(i, n) \\\n for (int i = blockIdx.x * blockDim.x + threadIdx.x; \\\n i < (n); \\\n i += blockDim.x * gridDim.x)\n\n__global__ void ROIPoolForward(const int nthreads, const float* bottom_data,\n const float spatial_scale, const int height, const int width,\n const int channels, const int pooled_height, const int pooled_width,\n const float* bottom_rois, float* top_data, int* argmax_data)\n{\n CUDA_KERNEL_LOOP(index, nthreads)\n {\n // (n, c, ph, pw) is an element in the pooled output\n // int n = index;\n // int pw = n % pooled_width;\n // n /= pooled_width;\n // int ph = n % pooled_height;\n // n /= pooled_height;\n // int c = n % channels;\n // n /= channels;\n int pw = index % pooled_width;\n int ph = (index / pooled_width) % pooled_height;\n int c = (index / pooled_width / pooled_height) % channels;\n int n = index / pooled_width / pooled_height / channels;\n\n // bottom_rois += n * 5;\n int roi_batch_ind = bottom_rois[n * 5 + 0];\n int roi_start_w = round(bottom_rois[n * 5 + 1] * spatial_scale);\n int roi_start_h = round(bottom_rois[n * 5 + 2] * spatial_scale);\n int roi_end_w = round(bottom_rois[n * 5 + 3] * spatial_scale);\n int roi_end_h = round(bottom_rois[n * 5 + 4] * spatial_scale);\n\n // Force malformed ROIs to be 1x1\n int roi_width = fmaxf(roi_end_w - roi_start_w + 1, 1);\n int roi_height = fmaxf(roi_end_h - roi_start_h + 1, 1);\n float bin_size_h = (float)(roi_height) / (float)(pooled_height);\n float bin_size_w = (float)(roi_width) / (float)(pooled_width);\n\n int hstart = (int)(floor((float)(ph) * bin_size_h));\n int wstart = (int)(floor((float)(pw) * bin_size_w));\n int hend = (int)(ceil((float)(ph + 1) * bin_size_h));\n int wend = (int)(ceil((float)(pw + 1) * bin_size_w));\n\n // Add roi offsets and clip to input boundaries\n hstart = fminf(fmaxf(hstart + roi_start_h, 0), height);\n hend = fminf(fmaxf(hend + roi_start_h, 0), height);\n wstart = fminf(fmaxf(wstart + roi_start_w, 0), width);\n wend = fminf(fmaxf(wend + roi_start_w, 0), width);\n bool is_empty = (hend <= hstart) || (wend <= wstart);\n\n // Define an empty pooling region to be zero\n float maxval = is_empty ? 0 : -FLT_MAX;\n // If nothing is pooled, argmax = -1 causes nothing to be backprop'd\n int maxidx = -1;\n // bottom_data += roi_batch_ind * channels * height * width;\n\n int bottom_data_batch_offset = roi_batch_ind * channels * height * width;\n int bottom_data_offset = bottom_data_batch_offset + c * height * width;\n\n for (int h = hstart; h < hend; ++h) {\n for (int w = wstart; w < wend; ++w) {\n // int bottom_index = (h * width + w) * channels + c;\n // int bottom_index = (c * height + h) * width + w;\n int bottom_index = h * width + w;\n if (bottom_data[bottom_data_offset + bottom_index] > maxval) {\n maxval = bottom_data[bottom_data_offset + bottom_index];\n maxidx = bottom_data_offset + bottom_index;\n }\n }\n }\n top_data[index] = maxval;\n if (argmax_data != NULL)\n argmax_data[index] = maxidx;\n }\n}\n\nint ROIPoolForwardLaucher(\n const float* bottom_data, const float spatial_scale, const int num_rois, const int height,\n const int width, const int channels, const int pooled_height,\n const int pooled_width, const float* bottom_rois,\n float* top_data, int* argmax_data, cudaStream_t stream)\n{\n const int kThreadsPerBlock = 1024;\n int output_size = num_rois * pooled_height * pooled_width * channels;\n cudaError_t err;\n\n ROIPoolForward<<<(output_size + kThreadsPerBlock - 1) / kThreadsPerBlock, kThreadsPerBlock, 0, stream>>>(\n output_size, bottom_data, spatial_scale, height, width, channels, pooled_height,\n pooled_width, bottom_rois, top_data, argmax_data);\n\n // dim3 blocks(DIVUP(output_size, kThreadsPerBlock),\n // DIVUP(output_size, kThreadsPerBlock));\n // dim3 threads(kThreadsPerBlock);\n //\n // ROIPoolForward<<>>(\n // output_size, bottom_data, spatial_scale, height, width, channels, pooled_height,\n // pooled_width, bottom_rois, top_data, argmax_data);\n\n err = cudaGetLastError();\n if(cudaSuccess != err)\n {\n fprintf( stderr, \"cudaCheckError() failed : %s\\n\", cudaGetErrorString( err ) );\n exit( -1 );\n }\n\n return 1;\n}\n\n\n__global__ void ROIPoolBackward(const int nthreads, const float* top_diff,\n const int* argmax_data, const int num_rois, const float spatial_scale,\n const int height, const int width, const int channels,\n const int pooled_height, const int pooled_width, float* bottom_diff,\n const float* bottom_rois) {\n CUDA_1D_KERNEL_LOOP(index, nthreads)\n {\n\n // (n, c, ph, pw) is an element in the pooled output\n int n = index;\n int w = n % width;\n n /= width;\n int h = n % height;\n n /= height;\n int c = n % channels;\n n /= channels;\n\n float gradient = 0;\n // Accumulate gradient over all ROIs that pooled this element\n for (int roi_n = 0; roi_n < num_rois; ++roi_n)\n {\n const float* offset_bottom_rois = bottom_rois + roi_n * 5;\n int roi_batch_ind = offset_bottom_rois[0];\n // Skip if ROI's batch index doesn't match n\n if (n != roi_batch_ind) {\n continue;\n }\n\n int roi_start_w = round(offset_bottom_rois[1] * spatial_scale);\n int roi_start_h = round(offset_bottom_rois[2] * spatial_scale);\n int roi_end_w = round(offset_bottom_rois[3] * spatial_scale);\n int roi_end_h = round(offset_bottom_rois[4] * spatial_scale);\n\n // Skip if ROI doesn't include (h, w)\n const bool in_roi = (w >= roi_start_w && w <= roi_end_w &&\n h >= roi_start_h && h <= roi_end_h);\n if (!in_roi) {\n continue;\n }\n\n int offset = roi_n * pooled_height * pooled_width * channels;\n const float* offset_top_diff = top_diff + offset;\n const int* offset_argmax_data = argmax_data + offset;\n\n // Compute feasible set of pooled units that could have pooled\n // this bottom unit\n\n // Force malformed ROIs to be 1x1\n int roi_width = fmaxf(roi_end_w - roi_start_w + 1, 1);\n int roi_height = fmaxf(roi_end_h - roi_start_h + 1, 1);\n\n float bin_size_h = (float)(roi_height) / (float)(pooled_height);\n float bin_size_w = (float)(roi_width) / (float)(pooled_width);\n\n int phstart = floor((float)(h - roi_start_h) / bin_size_h);\n int phend = ceil((float)(h - roi_start_h + 1) / bin_size_h);\n int pwstart = floor((float)(w - roi_start_w) / bin_size_w);\n int pwend = ceil((float)(w - roi_start_w + 1) / bin_size_w);\n\n phstart = fminf(fmaxf(phstart, 0), pooled_height);\n phend = fminf(fmaxf(phend, 0), pooled_height);\n pwstart = fminf(fmaxf(pwstart, 0), pooled_width);\n pwend = fminf(fmaxf(pwend, 0), pooled_width);\n\n for (int ph = phstart; ph < phend; ++ph) {\n for (int pw = pwstart; pw < pwend; ++pw) {\n if (offset_argmax_data[(c * pooled_height + ph) * pooled_width + pw] == index)\n {\n gradient += offset_top_diff[(c * pooled_height + ph) * pooled_width + pw];\n }\n }\n }\n }\n bottom_diff[index] = gradient;\n }\n}\n\nint ROIPoolBackwardLaucher(const float* top_diff, const float spatial_scale, const int batch_size, const int num_rois,\n const int height, const int width, const int channels, const int pooled_height,\n const int pooled_width, const float* bottom_rois,\n float* bottom_diff, const int* argmax_data, cudaStream_t stream)\n{\n const int kThreadsPerBlock = 1024;\n int output_size = batch_size * height * width * channels;\n cudaError_t err;\n\n ROIPoolBackward<<<(output_size + kThreadsPerBlock - 1) / kThreadsPerBlock, kThreadsPerBlock, 0, stream>>>(\n output_size, top_diff, argmax_data, num_rois, spatial_scale, height, width, channels, pooled_height,\n pooled_width, bottom_diff, bottom_rois);\n\n // dim3 blocks(DIVUP(output_size, kThreadsPerBlock),\n // DIVUP(output_size, kThreadsPerBlock));\n // dim3 threads(kThreadsPerBlock);\n //\n // ROIPoolBackward<<>>(\n // output_size, top_diff, argmax_data, num_rois, spatial_scale, height, width, channels, pooled_height,\n // pooled_width, bottom_diff, bottom_rois);\n\n err = cudaGetLastError();\n if(cudaSuccess != err)\n {\n fprintf( stderr, \"cudaCheckError() failed : %s\\n\", cudaGetErrorString( err ) );\n exit( -1 );\n }\n\n return 1;\n}\n\n\n// #ifdef __cplusplus\n// }\n// #endif\n"
},
{
"path": "lib/model/roi_pooling/src/roi_pooling_kernel.h",
"content": "#ifndef _ROI_POOLING_KERNEL\n#define _ROI_POOLING_KERNEL\n\n#ifdef __cplusplus\nextern \"C\" {\n#endif\n\nint ROIPoolForwardLaucher(\n const float* bottom_data, const float spatial_scale, const int num_rois, const int height,\n const int width, const int channels, const int pooled_height,\n const int pooled_width, const float* bottom_rois,\n float* top_data, int* argmax_data, cudaStream_t stream);\n\n\nint ROIPoolBackwardLaucher(const float* top_diff, const float spatial_scale, const int batch_size, const int num_rois,\n const int height, const int width, const int channels, const int pooled_height,\n const int pooled_width, const float* bottom_rois,\n float* bottom_diff, const int* argmax_data, cudaStream_t stream);\n\n#ifdef __cplusplus\n}\n#endif\n\n#endif\n\n"
},
{
"path": "lib/model/rpn/__init__.py",
"content": ""
},
{
"path": "lib/model/rpn/anchor_target_layer.py",
"content": "from __future__ import absolute_import\n# --------------------------------------------------------\n# Faster R-CNN\n# Copyright (c) 2015 Microsoft\n# Licensed under The MIT License [see LICENSE for details]\n# Written by Ross Girshick and Sean Bell\n# --------------------------------------------------------\n# --------------------------------------------------------\n# Reorganized and modified by Jianwei Yang and Jiasen Lu\n# --------------------------------------------------------\n\nimport torch\nimport torch.nn as nn\nimport numpy as np\nimport numpy.random as npr\n\nfrom model.utils.config import cfg\nfrom .generate_anchors import generate_anchors\nfrom .bbox_transform import clip_boxes, bbox_overlaps_batch, bbox_transform_batch\n\nimport pdb\n\nDEBUG = False\n\ntry:\n long # Python 2\nexcept NameError:\n long = int # Python 3\n\n\nclass _AnchorTargetLayer(nn.Module):\n \"\"\"\n Assign anchors to ground-truth targets. Produces anchor classification\n labels and bounding-box regression targets.\n \"\"\"\n def __init__(self, feat_stride, scales, ratios):\n super(_AnchorTargetLayer, self).__init__()\n\n self._feat_stride = feat_stride\n self._scales = scales\n anchor_scales = scales\n self._anchors = torch.from_numpy(generate_anchors(scales=np.array(anchor_scales), ratios=np.array(ratios))).float()\n self._num_anchors = self._anchors.size(0)\n\n # allow boxes to sit over the edge by a small amount\n self._allowed_border = 0 # default is 0\n\n def forward(self, input):\n # Algorithm:\n #\n # for each (H, W) location i\n # generate 9 anchor boxes centered on cell i\n # apply predicted bbox deltas at cell i to each of the 9 anchors\n # filter out-of-image anchors\n\n rpn_cls_score = input[0]\n gt_boxes = input[1]\n im_info = input[2]\n num_boxes = input[3]\n\n # map of shape (..., H, W)\n height, width = rpn_cls_score.size(2), rpn_cls_score.size(3)\n\n batch_size = gt_boxes.size(0)\n\n feat_height, feat_width = rpn_cls_score.size(2), rpn_cls_score.size(3)\n shift_x = np.arange(0, feat_width) * self._feat_stride\n shift_y = np.arange(0, feat_height) * self._feat_stride\n shift_x, shift_y = np.meshgrid(shift_x, shift_y)\n shifts = torch.from_numpy(np.vstack((shift_x.ravel(), shift_y.ravel(),\n shift_x.ravel(), shift_y.ravel())).transpose())\n shifts = shifts.contiguous().type_as(rpn_cls_score).float()\n\n A = self._num_anchors\n K = shifts.size(0)\n\n self._anchors = self._anchors.type_as(gt_boxes) # move to specific gpu.\n all_anchors = self._anchors.view(1, A, 4) + shifts.view(K, 1, 4)\n all_anchors = all_anchors.view(K * A, 4)\n\n total_anchors = int(K * A)\n\n keep = ((all_anchors[:, 0] >= -self._allowed_border) &\n (all_anchors[:, 1] >= -self._allowed_border) &\n (all_anchors[:, 2] < long(im_info[0][1]) + self._allowed_border) &\n (all_anchors[:, 3] < long(im_info[0][0]) + self._allowed_border))\n\n inds_inside = torch.nonzero(keep).view(-1)\n\n # keep only inside anchors\n anchors = all_anchors[inds_inside, :]\n\n # label: 1 is positive, 0 is negative, -1 is dont care\n labels = gt_boxes.new(batch_size, inds_inside.size(0)).fill_(-1)\n bbox_inside_weights = gt_boxes.new(batch_size, inds_inside.size(0)).zero_()\n bbox_outside_weights = gt_boxes.new(batch_size, inds_inside.size(0)).zero_()\n\n overlaps = bbox_overlaps_batch(anchors, gt_boxes)\n\n max_overlaps, argmax_overlaps = torch.max(overlaps, 2)\n gt_max_overlaps, _ = torch.max(overlaps, 1)\n\n if not cfg.TRAIN.RPN_CLOBBER_POSITIVES:\n labels[max_overlaps < cfg.TRAIN.RPN_NEGATIVE_OVERLAP] = 0\n\n gt_max_overlaps[gt_max_overlaps==0] = 1e-5\n keep = torch.sum(overlaps.eq(gt_max_overlaps.view(batch_size,1,-1).expand_as(overlaps)), 2)\n\n if torch.sum(keep) > 0:\n labels[keep>0] = 1\n\n # fg label: above threshold IOU\n labels[max_overlaps >= cfg.TRAIN.RPN_POSITIVE_OVERLAP] = 1\n\n if cfg.TRAIN.RPN_CLOBBER_POSITIVES:\n labels[max_overlaps < cfg.TRAIN.RPN_NEGATIVE_OVERLAP] = 0\n\n num_fg = int(cfg.TRAIN.RPN_FG_FRACTION * cfg.TRAIN.RPN_BATCHSIZE)\n\n sum_fg = torch.sum((labels == 1).int(), 1)\n sum_bg = torch.sum((labels == 0).int(), 1)\n\n for i in range(batch_size):\n # subsample positive labels if we have too many\n if sum_fg[i] > num_fg:\n fg_inds = torch.nonzero(labels[i] == 1).view(-1)\n # torch.randperm seems has a bug on multi-gpu setting that cause the segfault.\n # See https://github.com/pytorch/pytorch/issues/1868 for more details.\n # use numpy instead.\n #rand_num = torch.randperm(fg_inds.size(0)).type_as(gt_boxes).long()\n rand_num = torch.from_numpy(np.random.permutation(fg_inds.size(0))).type_as(gt_boxes).long()\n disable_inds = fg_inds[rand_num[:fg_inds.size(0)-num_fg]]\n labels[i][disable_inds] = -1\n\n num_bg = cfg.TRAIN.RPN_BATCHSIZE - sum_fg[i]\n\n # subsample negative labels if we have too many\n if sum_bg[i] > num_bg:\n bg_inds = torch.nonzero(labels[i] == 0).view(-1)\n #rand_num = torch.randperm(bg_inds.size(0)).type_as(gt_boxes).long()\n\n rand_num = torch.from_numpy(np.random.permutation(bg_inds.size(0))).type_as(gt_boxes).long()\n disable_inds = bg_inds[rand_num[:bg_inds.size(0)-num_bg]]\n labels[i][disable_inds] = -1\n\n offset = torch.arange(0, batch_size)*gt_boxes.size(1)\n\n argmax_overlaps = argmax_overlaps + offset.view(batch_size, 1).type_as(argmax_overlaps)\n bbox_targets = _compute_targets_batch(anchors, gt_boxes.view(-1,5)[argmax_overlaps.view(-1), :].view(batch_size, -1, 5))\n\n # use a single value instead of 4 values for easy index.\n bbox_inside_weights[labels==1] = cfg.TRAIN.RPN_BBOX_INSIDE_WEIGHTS[0]\n\n if cfg.TRAIN.RPN_POSITIVE_WEIGHT < 0:\n num_examples = torch.sum(labels[i] >= 0)\n positive_weights = 1.0 / num_examples\n negative_weights = 1.0 / num_examples\n else:\n assert ((cfg.TRAIN.RPN_POSITIVE_WEIGHT > 0) &\n (cfg.TRAIN.RPN_POSITIVE_WEIGHT < 1))\n\n bbox_outside_weights[labels == 1] = positive_weights\n bbox_outside_weights[labels == 0] = negative_weights\n\n labels = _unmap(labels, total_anchors, inds_inside, batch_size, fill=-1)\n bbox_targets = _unmap(bbox_targets, total_anchors, inds_inside, batch_size, fill=0)\n bbox_inside_weights = _unmap(bbox_inside_weights, total_anchors, inds_inside, batch_size, fill=0)\n bbox_outside_weights = _unmap(bbox_outside_weights, total_anchors, inds_inside, batch_size, fill=0)\n\n outputs = []\n\n labels = labels.view(batch_size, height, width, A).permute(0,3,1,2).contiguous()\n labels = labels.view(batch_size, 1, A * height, width)\n outputs.append(labels)\n\n bbox_targets = bbox_targets.view(batch_size, height, width, A*4).permute(0,3,1,2).contiguous()\n outputs.append(bbox_targets)\n\n anchors_count = bbox_inside_weights.size(1)\n bbox_inside_weights = bbox_inside_weights.view(batch_size,anchors_count,1).expand(batch_size, anchors_count, 4)\n\n bbox_inside_weights = bbox_inside_weights.contiguous().view(batch_size, height, width, 4*A)\\\n .permute(0,3,1,2).contiguous()\n\n outputs.append(bbox_inside_weights)\n\n bbox_outside_weights = bbox_outside_weights.view(batch_size,anchors_count,1).expand(batch_size, anchors_count, 4)\n bbox_outside_weights = bbox_outside_weights.contiguous().view(batch_size, height, width, 4*A)\\\n .permute(0,3,1,2).contiguous()\n outputs.append(bbox_outside_weights)\n\n return outputs\n\n def backward(self, top, propagate_down, bottom):\n \"\"\"This layer does not propagate gradients.\"\"\"\n pass\n\n def reshape(self, bottom, top):\n \"\"\"Reshaping happens during the call to forward.\"\"\"\n pass\n\ndef _unmap(data, count, inds, batch_size, fill=0):\n \"\"\" Unmap a subset of item (data) back to the original set of items (of\n size count) \"\"\"\n\n if data.dim() == 2:\n ret = torch.Tensor(batch_size, count).fill_(fill).type_as(data)\n ret[:, inds] = data\n else:\n ret = torch.Tensor(batch_size, count, data.size(2)).fill_(fill).type_as(data)\n ret[:, inds,:] = data\n return ret\n\n\ndef _compute_targets_batch(ex_rois, gt_rois):\n \"\"\"Compute bounding-box regression targets for an image.\"\"\"\n\n return bbox_transform_batch(ex_rois, gt_rois[:, :, :4])\n"
},
{
"path": "lib/model/rpn/bbox_transform.py",
"content": "# --------------------------------------------------------\n# Fast R-CNN\n# Copyright (c) 2015 Microsoft\n# Licensed under The MIT License [see LICENSE for details]\n# Written by Ross Girshick\n# --------------------------------------------------------\n# --------------------------------------------------------\n# Reorganized and modified by Jianwei Yang and Jiasen Lu\n# --------------------------------------------------------\n\nimport torch\nimport numpy as np\nimport pdb\n\ndef bbox_transform(ex_rois, gt_rois):\n ex_widths = ex_rois[:, 2] - ex_rois[:, 0] + 1.0\n ex_heights = ex_rois[:, 3] - ex_rois[:, 1] + 1.0\n ex_ctr_x = ex_rois[:, 0] + 0.5 * ex_widths\n ex_ctr_y = ex_rois[:, 1] + 0.5 * ex_heights\n\n gt_widths = gt_rois[:, 2] - gt_rois[:, 0] + 1.0\n gt_heights = gt_rois[:, 3] - gt_rois[:, 1] + 1.0\n gt_ctr_x = gt_rois[:, 0] + 0.5 * gt_widths\n gt_ctr_y = gt_rois[:, 1] + 0.5 * gt_heights\n\n targets_dx = (gt_ctr_x - ex_ctr_x) / ex_widths\n targets_dy = (gt_ctr_y - ex_ctr_y) / ex_heights\n targets_dw = torch.log(gt_widths / ex_widths)\n targets_dh = torch.log(gt_heights / ex_heights)\n\n targets = torch.stack(\n (targets_dx, targets_dy, targets_dw, targets_dh),1)\n\n return targets\n\ndef bbox_transform_batch(ex_rois, gt_rois):\n\n if ex_rois.dim() == 2:\n ex_widths = ex_rois[:, 2] - ex_rois[:, 0] + 1.0\n ex_heights = ex_rois[:, 3] - ex_rois[:, 1] + 1.0\n ex_ctr_x = ex_rois[:, 0] + 0.5 * ex_widths\n ex_ctr_y = ex_rois[:, 1] + 0.5 * ex_heights\n\n gt_widths = gt_rois[:, :, 2] - gt_rois[:, :, 0] + 1.0\n gt_heights = gt_rois[:, :, 3] - gt_rois[:, :, 1] + 1.0\n gt_ctr_x = gt_rois[:, :, 0] + 0.5 * gt_widths\n gt_ctr_y = gt_rois[:, :, 1] + 0.5 * gt_heights\n\n targets_dx = (gt_ctr_x - ex_ctr_x.view(1,-1).expand_as(gt_ctr_x)) / ex_widths\n targets_dy = (gt_ctr_y - ex_ctr_y.view(1,-1).expand_as(gt_ctr_y)) / ex_heights\n targets_dw = torch.log(gt_widths / ex_widths.view(1,-1).expand_as(gt_widths))\n targets_dh = torch.log(gt_heights / ex_heights.view(1,-1).expand_as(gt_heights))\n\n elif ex_rois.dim() == 3:\n ex_widths = ex_rois[:, :, 2] - ex_rois[:, :, 0] + 1.0\n ex_heights = ex_rois[:,:, 3] - ex_rois[:,:, 1] + 1.0\n ex_ctr_x = ex_rois[:, :, 0] + 0.5 * ex_widths\n ex_ctr_y = ex_rois[:, :, 1] + 0.5 * ex_heights\n\n gt_widths = gt_rois[:, :, 2] - gt_rois[:, :, 0] + 1.0\n gt_heights = gt_rois[:, :, 3] - gt_rois[:, :, 1] + 1.0\n gt_ctr_x = gt_rois[:, :, 0] + 0.5 * gt_widths\n gt_ctr_y = gt_rois[:, :, 1] + 0.5 * gt_heights\n\n targets_dx = (gt_ctr_x - ex_ctr_x) / ex_widths\n targets_dy = (gt_ctr_y - ex_ctr_y) / ex_heights\n targets_dw = torch.log(gt_widths / ex_widths)\n targets_dh = torch.log(gt_heights / ex_heights)\n else:\n raise ValueError('ex_roi input dimension is not correct.')\n\n targets = torch.stack(\n (targets_dx, targets_dy, targets_dw, targets_dh),2)\n\n return targets\n\ndef bbox_transform_inv(boxes, deltas, batch_size):\n widths = boxes[:, :, 2] - boxes[:, :, 0] + 1.0\n heights = boxes[:, :, 3] - boxes[:, :, 1] + 1.0\n ctr_x = boxes[:, :, 0] + 0.5 * widths\n ctr_y = boxes[:, :, 1] + 0.5 * heights\n\n dx = deltas[:, :, 0::4]\n dy = deltas[:, :, 1::4]\n dw = deltas[:, :, 2::4]\n dh = deltas[:, :, 3::4]\n\n pred_ctr_x = dx * widths.unsqueeze(2) + ctr_x.unsqueeze(2)\n pred_ctr_y = dy * heights.unsqueeze(2) + ctr_y.unsqueeze(2)\n pred_w = torch.exp(dw) * widths.unsqueeze(2)\n pred_h = torch.exp(dh) * heights.unsqueeze(2)\n\n pred_boxes = deltas.clone()\n # x1\n pred_boxes[:, :, 0::4] = pred_ctr_x - 0.5 * pred_w\n # y1\n pred_boxes[:, :, 1::4] = pred_ctr_y - 0.5 * pred_h\n # x2\n pred_boxes[:, :, 2::4] = pred_ctr_x + 0.5 * pred_w\n # y2\n pred_boxes[:, :, 3::4] = pred_ctr_y + 0.5 * pred_h\n\n return pred_boxes\n\ndef clip_boxes_batch(boxes, im_shape, batch_size):\n \"\"\"\n Clip boxes to image boundaries.\n \"\"\"\n num_rois = boxes.size(1)\n\n boxes[boxes < 0] = 0\n # batch_x = (im_shape[:,0]-1).view(batch_size, 1).expand(batch_size, num_rois)\n # batch_y = (im_shape[:,1]-1).view(batch_size, 1).expand(batch_size, num_rois)\n\n batch_x = im_shape[:, 1] - 1\n batch_y = im_shape[:, 0] - 1\n\n boxes[:,:,0][boxes[:,:,0] > batch_x] = batch_x\n boxes[:,:,1][boxes[:,:,1] > batch_y] = batch_y\n boxes[:,:,2][boxes[:,:,2] > batch_x] = batch_x\n boxes[:,:,3][boxes[:,:,3] > batch_y] = batch_y\n\n return boxes\n\ndef clip_boxes(boxes, im_shape, batch_size):\n\n for i in range(batch_size):\n boxes[i,:,0::4].clamp_(0, im_shape[i, 1]-1)\n boxes[i,:,1::4].clamp_(0, im_shape[i, 0]-1)\n boxes[i,:,2::4].clamp_(0, im_shape[i, 1]-1)\n boxes[i,:,3::4].clamp_(0, im_shape[i, 0]-1)\n\n return boxes\n\n\ndef bbox_overlaps(anchors, gt_boxes):\n \"\"\"\n anchors: (N, 4) ndarray of float\n gt_boxes: (K, 4) ndarray of float\n\n overlaps: (N, K) ndarray of overlap between boxes and query_boxes\n \"\"\"\n N = anchors.size(0)\n K = gt_boxes.size(0)\n\n gt_boxes_area = ((gt_boxes[:,2] - gt_boxes[:,0] + 1) *\n (gt_boxes[:,3] - gt_boxes[:,1] + 1)).view(1, K)\n\n anchors_area = ((anchors[:,2] - anchors[:,0] + 1) *\n (anchors[:,3] - anchors[:,1] + 1)).view(N, 1)\n\n boxes = anchors.view(N, 1, 4).expand(N, K, 4)\n query_boxes = gt_boxes.view(1, K, 4).expand(N, K, 4)\n\n iw = (torch.min(boxes[:,:,2], query_boxes[:,:,2]) -\n torch.max(boxes[:,:,0], query_boxes[:,:,0]) + 1)\n iw[iw < 0] = 0\n\n ih = (torch.min(boxes[:,:,3], query_boxes[:,:,3]) -\n torch.max(boxes[:,:,1], query_boxes[:,:,1]) + 1)\n ih[ih < 0] = 0\n\n ua = anchors_area + gt_boxes_area - (iw * ih)\n overlaps = iw * ih / ua\n\n return overlaps\n\ndef bbox_overlaps_batch(anchors, gt_boxes):\n \"\"\"\n anchors: (N, 4) ndarray of float\n gt_boxes: (b, K, 5) ndarray of float\n\n overlaps: (N, K) ndarray of overlap between boxes and query_boxes\n \"\"\"\n batch_size = gt_boxes.size(0)\n\n\n if anchors.dim() == 2:\n\n N = anchors.size(0)\n K = gt_boxes.size(1)\n\n anchors = anchors.view(1, N, 4).expand(batch_size, N, 4).contiguous()\n gt_boxes = gt_boxes[:,:,:4].contiguous()\n\n\n gt_boxes_x = (gt_boxes[:,:,2] - gt_boxes[:,:,0] + 1)\n gt_boxes_y = (gt_boxes[:,:,3] - gt_boxes[:,:,1] + 1)\n gt_boxes_area = (gt_boxes_x * gt_boxes_y).view(batch_size, 1, K)\n\n anchors_boxes_x = (anchors[:,:,2] - anchors[:,:,0] + 1)\n anchors_boxes_y = (anchors[:,:,3] - anchors[:,:,1] + 1)\n anchors_area = (anchors_boxes_x * anchors_boxes_y).view(batch_size, N, 1)\n\n gt_area_zero = (gt_boxes_x == 1) & (gt_boxes_y == 1)\n anchors_area_zero = (anchors_boxes_x == 1) & (anchors_boxes_y == 1)\n\n boxes = anchors.view(batch_size, N, 1, 4).expand(batch_size, N, K, 4)\n query_boxes = gt_boxes.view(batch_size, 1, K, 4).expand(batch_size, N, K, 4)\n\n iw = (torch.min(boxes[:,:,:,2], query_boxes[:,:,:,2]) -\n torch.max(boxes[:,:,:,0], query_boxes[:,:,:,0]) + 1)\n iw[iw < 0] = 0\n\n ih = (torch.min(boxes[:,:,:,3], query_boxes[:,:,:,3]) -\n torch.max(boxes[:,:,:,1], query_boxes[:,:,:,1]) + 1)\n ih[ih < 0] = 0\n ua = anchors_area + gt_boxes_area - (iw * ih)\n overlaps = iw * ih / ua\n\n # mask the overlap here.\n overlaps.masked_fill_(gt_area_zero.view(batch_size, 1, K).expand(batch_size, N, K), 0)\n overlaps.masked_fill_(anchors_area_zero.view(batch_size, N, 1).expand(batch_size, N, K), -1)\n\n elif anchors.dim() == 3:\n N = anchors.size(1)\n K = gt_boxes.size(1)\n\n if anchors.size(2) == 4:\n anchors = anchors[:,:,:4].contiguous()\n else:\n anchors = anchors[:,:,1:5].contiguous()\n\n gt_boxes = gt_boxes[:,:,:4].contiguous()\n\n gt_boxes_x = (gt_boxes[:,:,2] - gt_boxes[:,:,0] + 1)\n gt_boxes_y = (gt_boxes[:,:,3] - gt_boxes[:,:,1] + 1)\n gt_boxes_area = (gt_boxes_x * gt_boxes_y).view(batch_size, 1, K)\n\n anchors_boxes_x = (anchors[:,:,2] - anchors[:,:,0] + 1)\n anchors_boxes_y = (anchors[:,:,3] - anchors[:,:,1] + 1)\n anchors_area = (anchors_boxes_x * anchors_boxes_y).view(batch_size, N, 1)\n\n gt_area_zero = (gt_boxes_x == 1) & (gt_boxes_y == 1)\n anchors_area_zero = (anchors_boxes_x == 1) & (anchors_boxes_y == 1)\n\n boxes = anchors.view(batch_size, N, 1, 4).expand(batch_size, N, K, 4)\n query_boxes = gt_boxes.view(batch_size, 1, K, 4).expand(batch_size, N, K, 4)\n\n iw = (torch.min(boxes[:,:,:,2], query_boxes[:,:,:,2]) -\n torch.max(boxes[:,:,:,0], query_boxes[:,:,:,0]) + 1)\n iw[iw < 0] = 0\n\n ih = (torch.min(boxes[:,:,:,3], query_boxes[:,:,:,3]) -\n torch.max(boxes[:,:,:,1], query_boxes[:,:,:,1]) + 1)\n ih[ih < 0] = 0\n ua = anchors_area + gt_boxes_area - (iw * ih)\n\n overlaps = iw * ih / ua\n\n # mask the overlap here.\n overlaps.masked_fill_(gt_area_zero.view(batch_size, 1, K).expand(batch_size, N, K), 0)\n overlaps.masked_fill_(anchors_area_zero.view(batch_size, N, 1).expand(batch_size, N, K), -1)\n else:\n raise ValueError('anchors input dimension is not correct.')\n\n return overlaps\n"
},
{
"path": "lib/model/rpn/generate_anchors.py",
"content": "from __future__ import print_function\n# --------------------------------------------------------\n# Faster R-CNN\n# Copyright (c) 2015 Microsoft\n# Licensed under The MIT License [see LICENSE for details]\n# Written by Ross Girshick and Sean Bell\n# --------------------------------------------------------\n\nimport numpy as np\nimport pdb\n\n# Verify that we compute the same anchors as Shaoqing's matlab implementation:\n#\n# >> load output/rpn_cachedir/faster_rcnn_VOC2007_ZF_stage1_rpn/anchors.mat\n# >> anchors\n#\n# anchors =\n#\n# -83 -39 100 56\n# -175 -87 192 104\n# -359 -183 376 200\n# -55 -55 72 72\n# -119 -119 136 136\n# -247 -247 264 264\n# -35 -79 52 96\n# -79 -167 96 184\n# -167 -343 184 360\n\n#array([[ -83., -39., 100., 56.],\n# [-175., -87., 192., 104.],\n# [-359., -183., 376., 200.],\n# [ -55., -55., 72., 72.],\n# [-119., -119., 136., 136.],\n# [-247., -247., 264., 264.],\n# [ -35., -79., 52., 96.],\n# [ -79., -167., 96., 184.],\n# [-167., -343., 184., 360.]])\n\ntry:\n xrange # Python 2\nexcept NameError:\n xrange = range # Python 3\n\n\ndef generate_anchors(base_size=16, ratios=[0.5, 1, 2],\n scales=2**np.arange(3, 6)):\n \"\"\"\n Generate anchor (reference) windows by enumerating aspect ratios X\n scales wrt a reference (0, 0, 15, 15) window.\n \"\"\"\n\n base_anchor = np.array([1, 1, base_size, base_size]) - 1\n ratio_anchors = _ratio_enum(base_anchor, ratios)\n anchors = np.vstack([_scale_enum(ratio_anchors[i, :], scales)\n for i in xrange(ratio_anchors.shape[0])])\n return anchors\n\ndef _whctrs(anchor):\n \"\"\"\n Return width, height, x center, and y center for an anchor (window).\n \"\"\"\n\n w = anchor[2] - anchor[0] + 1\n h = anchor[3] - anchor[1] + 1\n x_ctr = anchor[0] + 0.5 * (w - 1)\n y_ctr = anchor[1] + 0.5 * (h - 1)\n return w, h, x_ctr, y_ctr\n\ndef _mkanchors(ws, hs, x_ctr, y_ctr):\n \"\"\"\n Given a vector of widths (ws) and heights (hs) around a center\n (x_ctr, y_ctr), output a set of anchors (windows).\n \"\"\"\n\n ws = ws[:, np.newaxis]\n hs = hs[:, np.newaxis]\n anchors = np.hstack((x_ctr - 0.5 * (ws - 1),\n y_ctr - 0.5 * (hs - 1),\n x_ctr + 0.5 * (ws - 1),\n y_ctr + 0.5 * (hs - 1)))\n return anchors\n\ndef _ratio_enum(anchor, ratios):\n \"\"\"\n Enumerate a set of anchors for each aspect ratio wrt an anchor.\n \"\"\"\n\n w, h, x_ctr, y_ctr = _whctrs(anchor)\n size = w * h\n size_ratios = size / ratios\n ws = np.round(np.sqrt(size_ratios))\n hs = np.round(ws * ratios)\n anchors = _mkanchors(ws, hs, x_ctr, y_ctr)\n return anchors\n\ndef _scale_enum(anchor, scales):\n \"\"\"\n Enumerate a set of anchors for each scale wrt an anchor.\n \"\"\"\n\n w, h, x_ctr, y_ctr = _whctrs(anchor)\n ws = w * scales\n hs = h * scales\n anchors = _mkanchors(ws, hs, x_ctr, y_ctr)\n return anchors\n\nif __name__ == '__main__':\n import time\n t = time.time()\n a = generate_anchors()\n print(time.time() - t)\n print(a)\n from IPython import embed; embed()\n"
},
{
"path": "lib/model/rpn/proposal_layer.py",
"content": "from __future__ import absolute_import\n# --------------------------------------------------------\n# Faster R-CNN\n# Copyright (c) 2015 Microsoft\n# Licensed under The MIT License [see LICENSE for details]\n# Written by Ross Girshick and Sean Bell\n# --------------------------------------------------------\n# --------------------------------------------------------\n# Reorganized and modified by Jianwei Yang and Jiasen Lu\n# --------------------------------------------------------\n\nimport torch\nimport torch.nn as nn\nimport numpy as np\nimport math\nimport yaml\nfrom model.utils.config import cfg\nfrom .generate_anchors import generate_anchors\nfrom .bbox_transform import bbox_transform_inv, clip_boxes, clip_boxes_batch\nfrom model.nms.nms_wrapper import nms\n\nimport pdb\n\nDEBUG = False\n\nclass _ProposalLayer(nn.Module):\n \"\"\"\n Outputs object detection proposals by applying estimated bounding-box\n transformations to a set of regular boxes (called \"anchors\").\n \"\"\"\n\n def __init__(self, feat_stride, scales, ratios):\n super(_ProposalLayer, self).__init__()\n\n self._feat_stride = feat_stride\n self._anchors = torch.from_numpy(generate_anchors(scales=np.array(scales), \n ratios=np.array(ratios))).float()\n self._num_anchors = self._anchors.size(0)\n\n # rois blob: holds R regions of interest, each is a 5-tuple\n # (n, x1, y1, x2, y2) specifying an image batch index n and a\n # rectangle (x1, y1, x2, y2)\n # top[0].reshape(1, 5)\n #\n # # scores blob: holds scores for R regions of interest\n # if len(top) > 1:\n # top[1].reshape(1, 1, 1, 1)\n\n def forward(self, input):\n\n # Algorithm:\n #\n # for each (H, W) location i\n # generate A anchor boxes centered on cell i\n # apply predicted bbox deltas at cell i to each of the A anchors\n # clip predicted boxes to image\n # remove predicted boxes with either height or width < threshold\n # sort all (proposal, score) pairs by score from highest to lowest\n # take top pre_nms_topN proposals before NMS\n # apply NMS with threshold 0.7 to remaining proposals\n # take after_nms_topN proposals after NMS\n # return the top proposals (-> RoIs top, scores top)\n\n\n # the first set of _num_anchors channels are bg probs\n # the second set are the fg probs\n scores = input[0][:, self._num_anchors:, :, :]\n bbox_deltas = input[1]\n im_info = input[2]\n cfg_key = input[3]\n\n pre_nms_topN = cfg[cfg_key].RPN_PRE_NMS_TOP_N\n post_nms_topN = cfg[cfg_key].RPN_POST_NMS_TOP_N\n nms_thresh = cfg[cfg_key].RPN_NMS_THRESH\n min_size = cfg[cfg_key].RPN_MIN_SIZE\n\n batch_size = bbox_deltas.size(0)\n\n feat_height, feat_width = scores.size(2), scores.size(3)\n shift_x = np.arange(0, feat_width) * self._feat_stride\n shift_y = np.arange(0, feat_height) * self._feat_stride\n shift_x, shift_y = np.meshgrid(shift_x, shift_y)\n shifts = torch.from_numpy(np.vstack((shift_x.ravel(), shift_y.ravel(),\n shift_x.ravel(), shift_y.ravel())).transpose())\n shifts = shifts.contiguous().type_as(scores).float()\n\n A = self._num_anchors\n K = shifts.size(0)\n\n self._anchors = self._anchors.type_as(scores)\n # anchors = self._anchors.view(1, A, 4) + shifts.view(1, K, 4).permute(1, 0, 2).contiguous()\n anchors = self._anchors.view(1, A, 4) + shifts.view(K, 1, 4)\n anchors = anchors.view(1, K * A, 4).expand(batch_size, K * A, 4)\n\n # Transpose and reshape predicted bbox transformations to get them\n # into the same order as the anchors:\n\n bbox_deltas = bbox_deltas.permute(0, 2, 3, 1).contiguous()\n bbox_deltas = bbox_deltas.view(batch_size, -1, 4)\n\n # Same story for the scores:\n scores = scores.permute(0, 2, 3, 1).contiguous()\n scores = scores.view(batch_size, -1)\n\n # Convert anchors into proposals via bbox transformations\n proposals = bbox_transform_inv(anchors, bbox_deltas, batch_size)\n\n # 2. clip predicted boxes to image\n proposals = clip_boxes(proposals, im_info, batch_size)\n # proposals = clip_boxes_batch(proposals, im_info, batch_size)\n\n # assign the score to 0 if it's non keep.\n # keep = self._filter_boxes(proposals, min_size * im_info[:, 2])\n\n # trim keep index to make it euqal over batch\n # keep_idx = torch.cat(tuple(keep_idx), 0)\n\n # scores_keep = scores.view(-1)[keep_idx].view(batch_size, trim_size)\n # proposals_keep = proposals.view(-1, 4)[keep_idx, :].contiguous().view(batch_size, trim_size, 4)\n \n # _, order = torch.sort(scores_keep, 1, True)\n \n scores_keep = scores\n proposals_keep = proposals\n _, order = torch.sort(scores_keep, 1, True)\n\n output = scores.new(batch_size, post_nms_topN, 5).zero_()\n for i in range(batch_size):\n # # 3. remove predicted boxes with either height or width < threshold\n # # (NOTE: convert min_size to input image scale stored in im_info[2])\n proposals_single = proposals_keep[i]\n scores_single = scores_keep[i]\n\n # # 4. sort all (proposal, score) pairs by score from highest to lowest\n # # 5. take top pre_nms_topN (e.g. 6000)\n order_single = order[i]\n\n if pre_nms_topN > 0 and pre_nms_topN < scores_keep.numel():\n order_single = order_single[:pre_nms_topN]\n\n proposals_single = proposals_single[order_single, :]\n scores_single = scores_single[order_single].view(-1,1)\n\n # 6. apply nms (e.g. threshold = 0.7)\n # 7. take after_nms_topN (e.g. 300)\n # 8. return the top proposals (-> RoIs top)\n\n keep_idx_i = nms(torch.cat((proposals_single, scores_single), 1), nms_thresh)\n keep_idx_i = keep_idx_i.long().view(-1)\n\n if post_nms_topN > 0:\n keep_idx_i = keep_idx_i[:post_nms_topN]\n proposals_single = proposals_single[keep_idx_i, :]\n scores_single = scores_single[keep_idx_i, :]\n\n # padding 0 at the end.\n num_proposal = proposals_single.size(0)\n output[i,:,0] = i\n output[i,:num_proposal,1:] = proposals_single\n\n return output\n\n def backward(self, top, propagate_down, bottom):\n \"\"\"This layer does not propagate gradients.\"\"\"\n pass\n\n def reshape(self, bottom, top):\n \"\"\"Reshaping happens during the call to forward.\"\"\"\n pass\n\n def _filter_boxes(self, boxes, min_size):\n \"\"\"Remove all boxes with any side smaller than min_size.\"\"\"\n ws = boxes[:, :, 2] - boxes[:, :, 0] + 1\n hs = boxes[:, :, 3] - boxes[:, :, 1] + 1\n keep = ((ws >= min_size.view(-1,1).expand_as(ws)) & (hs >= min_size.view(-1,1).expand_as(hs)))\n return keep\n"
},
{
"path": "lib/model/rpn/proposal_layer_region.py",
"content": "from __future__ import absolute_import\n# --------------------------------------------------------\n# Faster R-CNN\n# Copyright (c) 2015 Microsoft\n# Licensed under The MIT License [see LICENSE for details]\n# Written by Ross Girshick and Sean Bell\n# --------------------------------------------------------\n# --------------------------------------------------------\n# Reorganized and modified by Jianwei Yang and Jiasen Lu\n# --------------------------------------------------------\n\nimport torch\nimport torch.nn as nn\nimport numpy as np\nimport math\nimport yaml\nfrom model.utils.config import cfg\nfrom .generate_anchors import generate_anchors\nfrom .bbox_transform import bbox_transform_inv, clip_boxes, clip_boxes_batch\nfrom model.nms.nms_wrapper import nms\n\nimport pdb\n\nDEBUG = False\n\nclass _ProposalLayer(nn.Module):\n \"\"\"\n Outputs object detection proposals by applying estimated bounding-box\n transformations to a set of regular boxes (called \"anchors\").\n \"\"\"\n\n def __init__(self, feat_stride, scales, ratios):\n super(_ProposalLayer, self).__init__()\n\n self._feat_stride = feat_stride\n self._anchors = torch.from_numpy(generate_anchors(scales=np.array(scales), \n ratios=np.array(ratios))).float()\n self._num_anchors = self._anchors.size(0)\n\n # rois blob: holds R regions of interest, each is a 5-tuple\n # (n, x1, y1, x2, y2) specifying an image batch index n and a\n # rectangle (x1, y1, x2, y2)\n # top[0].reshape(1, 5)\n #\n # # scores blob: holds scores for R regions of interest\n # if len(top) > 1:\n # top[1].reshape(1, 1, 1, 1)\n\n def forward(self, input):\n\n # Algorithm:\n #\n # for each (H, W) location i\n # generate A anchor boxes centered on cell i\n # apply predicted bbox deltas at cell i to each of the A anchors\n # clip predicted boxes to image\n # remove predicted boxes with either height or width < threshold\n # sort all (proposal, score) pairs by score from highest to lowest\n # take top pre_nms_topN proposals before NMS\n # apply NMS with threshold 0.7 to remaining proposals\n # take after_nms_topN proposals after NMS\n # return the top proposals (-> RoIs top, scores top)\n\n\n # the first set of _num_anchors channels are bg probs\n # the second set are the fg probs\n scores = input[0][:, self._num_anchors:, :, :]\n bbox_deltas = input[1]\n im_info = input[2]\n cfg_key = input[3]\n\n pre_nms_topN = cfg[cfg_key].RPN_PRE_NMS_TOP_N\n post_nms_topN = cfg[cfg_key].RPN_POST_NMS_TOP_N\n nms_thresh = cfg[cfg_key].RPN_NMS_THRESH\n min_size = cfg[cfg_key].RPN_MIN_SIZE\n\n batch_size = bbox_deltas.size(0)\n\n feat_height, feat_width = scores.size(2), scores.size(3)\n shift_x = np.arange(0, feat_width) * self._feat_stride\n shift_y = np.arange(0, feat_height) * self._feat_stride\n shift_x, shift_y = np.meshgrid(shift_x, shift_y)\n shifts = torch.from_numpy(np.vstack((shift_x.ravel(), shift_y.ravel(),\n shift_x.ravel(), shift_y.ravel())).transpose())\n shifts = shifts.contiguous().type_as(scores).float()\n\n A = self._num_anchors\n K = shifts.size(0)\n\n self._anchors = self._anchors.type_as(scores)\n # anchors = self._anchors.view(1, A, 4) + shifts.view(1, K, 4).permute(1, 0, 2).contiguous()\n anchors = self._anchors.view(1, A, 4) + shifts.view(K, 1, 4)\n anchors = anchors.view(1, K * A, 4).expand(batch_size, K * A, 4)\n\n # Transpose and reshape predicted bbox transformations to get them\n # into the same order as the anchors:\n\n bbox_deltas = bbox_deltas.permute(0, 2, 3, 1).contiguous()\n bbox_deltas = bbox_deltas.view(batch_size, -1, 4)\n\n # Same story for the scores:\n scores = scores.permute(0, 2, 3, 1).contiguous()\n scores = scores.view(batch_size, -1)\n\n # Convert anchors into proposals via bbox transformations\n proposals = bbox_transform_inv(anchors, bbox_deltas, batch_size)\n\n # 2. clip predicted boxes to image\n proposals = clip_boxes(proposals, im_info, batch_size)\n # proposals = clip_boxes_batch(proposals, im_info, batch_size)\n\n # assign the score to 0 if it's non keep.\n # keep = self._filter_boxes(proposals, min_size * im_info[:, 2])\n\n # trim keep index to make it euqal over batch\n # keep_idx = torch.cat(tuple(keep_idx), 0)\n\n # scores_keep = scores.view(-1)[keep_idx].view(batch_size, trim_size)\n # proposals_keep = proposals.view(-1, 4)[keep_idx, :].contiguous().view(batch_size, trim_size, 4)\n \n # _, order = torch.sort(scores_keep, 1, True)\n \n scores_keep = scores\n proposals_keep = proposals\n _, order = torch.sort(scores_keep, 1, True)\n\n output = scores.new(batch_size, post_nms_topN, 5).zero_()\n output_cls_score = scores.new(batch_size, post_nms_topN, 2).zero_()\n for i in range(batch_size):\n # # 3. remove predicted boxes with either height or width < threshold\n # # (NOTE: convert min_size to input image scale stored in im_info[2])\n proposals_single = proposals_keep[i]\n scores_single = scores_keep[i]\n\n # # 4. sort all (proposal, score) pairs by score from highest to lowest\n # # 5. take top pre_nms_topN (e.g. 6000)\n order_single = order[i]\n\n if pre_nms_topN > 0 and pre_nms_topN < scores_keep.numel():\n order_single = order_single[:pre_nms_topN]\n\n proposals_single = proposals_single[order_single, :]\n scores_single = scores_single[order_single].view(-1,1)\n\n # 6. apply nms (e.g. threshold = 0.7)\n # 7. take after_nms_topN (e.g. 300)\n # 8. return the top proposals (-> RoIs top)\n\n keep_idx_i = nms(torch.cat((proposals_single, scores_single), 1), nms_thresh)\n keep_idx_i = keep_idx_i.long().view(-1)\n\n if post_nms_topN > 0:\n keep_idx_i = keep_idx_i[:post_nms_topN]\n proposals_single = proposals_single[keep_idx_i, :]\n scores_single = scores_single[keep_idx_i, :]\n\n # padding 0 at the end.\n num_proposal = proposals_single.size(0)\n output[i,:,0] = i\n output[i,:num_proposal,1:] = proposals_single\n output_cls_score[i,:,0] = i\n output_cls_score[i,:num_proposal,1] = scores_single\n\n return output, output_cls_score\n\n def backward(self, top, propagate_down, bottom):\n \"\"\"This layer does not propagate gradients.\"\"\"\n pass\n\n def reshape(self, bottom, top):\n \"\"\"Reshaping happens during the call to forward.\"\"\"\n pass\n\n def _filter_boxes(self, boxes, min_size):\n \"\"\"Remove all boxes with any side smaller than min_size.\"\"\"\n ws = boxes[:, :, 2] - boxes[:, :, 0] + 1\n hs = boxes[:, :, 3] - boxes[:, :, 1] + 1\n keep = ((ws >= min_size.view(-1,1).expand_as(ws)) & (hs >= min_size.view(-1,1).expand_as(hs)))\n return keep\n"
},
{
"path": "lib/model/rpn/proposal_target_layer_cascade.py",
"content": "from __future__ import absolute_import\n# --------------------------------------------------------\n# Faster R-CNN\n# Copyright (c) 2015 Microsoft\n# Licensed under The MIT License [see LICENSE for details]\n# Written by Ross Girshick and Sean Bell\n# --------------------------------------------------------\n# --------------------------------------------------------\n# Reorganized and modified by Jianwei Yang and Jiasen Lu\n# --------------------------------------------------------\n\nimport torch\nimport torch.nn as nn\nimport numpy as np\nimport numpy.random as npr\nfrom ..utils.config import cfg\nfrom .bbox_transform import bbox_overlaps_batch, bbox_transform_batch\nimport pdb\n\nclass _ProposalTargetLayer(nn.Module):\n \"\"\"\n Assign object detection proposals to ground-truth targets. Produces proposal\n classification labels and bounding-box regression targets.\n \"\"\"\n\n def __init__(self, nclasses):\n super(_ProposalTargetLayer, self).__init__()\n self._num_classes = nclasses\n self.BBOX_NORMALIZE_MEANS = torch.FloatTensor(cfg.TRAIN.BBOX_NORMALIZE_MEANS)\n self.BBOX_NORMALIZE_STDS = torch.FloatTensor(cfg.TRAIN.BBOX_NORMALIZE_STDS)\n self.BBOX_INSIDE_WEIGHTS = torch.FloatTensor(cfg.TRAIN.BBOX_INSIDE_WEIGHTS)\n\n def forward(self, all_rois, gt_boxes, num_boxes):\n\n self.BBOX_NORMALIZE_MEANS = self.BBOX_NORMALIZE_MEANS.type_as(gt_boxes)\n self.BBOX_NORMALIZE_STDS = self.BBOX_NORMALIZE_STDS.type_as(gt_boxes)\n self.BBOX_INSIDE_WEIGHTS = self.BBOX_INSIDE_WEIGHTS.type_as(gt_boxes)\n\n num_images = 1\n rois_per_image = int(cfg.TRAIN.BATCH_SIZE / num_images)\n fg_rois_per_image = int(np.round(cfg.TRAIN.FG_FRACTION * rois_per_image))\n fg_rois_per_image = 1 if fg_rois_per_image == 0 else fg_rois_per_image\n\n labels, rois, bbox_targets, bbox_inside_weights = self._sample_rois_pytorch(\n all_rois, gt_boxes, fg_rois_per_image,\n rois_per_image, self._num_classes)\n\n bbox_outside_weights = (bbox_inside_weights > 0).float()\n\n return rois, labels, bbox_targets, bbox_inside_weights, bbox_outside_weights\n\n def backward(self, top, propagate_down, bottom):\n \"\"\"This layer does not propagate gradients.\"\"\"\n pass\n\n def reshape(self, bottom, top):\n \"\"\"Reshaping happens during the call to forward.\"\"\"\n pass\n\n def _get_bbox_regression_labels_pytorch(self, bbox_target_data, labels_batch, num_classes):\n \"\"\"Bounding-box regression targets (bbox_target_data) are stored in a\n compact form b x N x (class, tx, ty, tw, th)\n\n This function expands those targets into the 4-of-4*K representation used\n by the network (i.e. only one class has non-zero targets).\n\n Returns:\n bbox_target (ndarray): b x N x 4K blob of regression targets\n bbox_inside_weights (ndarray): b x N x 4K blob of loss weights\n \"\"\"\n batch_size = labels_batch.size(0)\n rois_per_image = labels_batch.size(1)\n clss = labels_batch\n bbox_targets = bbox_target_data.new(batch_size, rois_per_image, 4).zero_()\n bbox_inside_weights = bbox_target_data.new(bbox_targets.size()).zero_()\n\n for b in range(batch_size):\n # assert clss[b].sum() > 0\n if clss[b].sum() == 0:\n continue\n inds = torch.nonzero(clss[b] > 0).view(-1)\n for i in range(inds.numel()):\n ind = inds[i]\n bbox_targets[b, ind, :] = bbox_target_data[b, ind, :]\n bbox_inside_weights[b, ind, :] = self.BBOX_INSIDE_WEIGHTS\n\n return bbox_targets, bbox_inside_weights\n\n\n def _compute_targets_pytorch(self, ex_rois, gt_rois):\n \"\"\"Compute bounding-box regression targets for an image.\"\"\"\n\n assert ex_rois.size(1) == gt_rois.size(1)\n assert ex_rois.size(2) == 4\n assert gt_rois.size(2) == 4\n\n batch_size = ex_rois.size(0)\n rois_per_image = ex_rois.size(1)\n\n targets = bbox_transform_batch(ex_rois, gt_rois)\n\n if cfg.TRAIN.BBOX_NORMALIZE_TARGETS_PRECOMPUTED:\n # Optionally normalize targets by a precomputed mean and stdev\n targets = ((targets - self.BBOX_NORMALIZE_MEANS.expand_as(targets))\n / self.BBOX_NORMALIZE_STDS.expand_as(targets))\n\n return targets\n\n\n def _sample_rois_pytorch(self, all_rois, gt_boxes, fg_rois_per_image, rois_per_image, num_classes):\n \"\"\"Generate a random sample of RoIs comprising foreground and background\n examples.\n \"\"\"\n # overlaps: (rois x gt_boxes)\n\n overlaps = bbox_overlaps_batch(all_rois, gt_boxes)\n # max_overlaps = max overlap of (candidate rois with gt_rois)\n max_overlaps, gt_assignment = torch.max(overlaps, 2)\n batch_size = overlaps.size(0)\n num_proposal = overlaps.size(1)\n num_boxes_per_img = overlaps.size(2)\n\n offset = torch.arange(0, batch_size)*gt_boxes.size(1)\n offset = offset.view(-1, 1).type_as(gt_assignment) + gt_assignment\n \n labels = gt_boxes[:,:,4].contiguous().view(-1).index(offset.view(-1))\\\n .view(batch_size, -1)\n\n labels_batch = labels.new(batch_size, rois_per_image).zero_()\n rois_batch = all_rois.new(batch_size, rois_per_image, 5).zero_() # get rois_per_image front of rois\n gt_rois_batch = all_rois.new(batch_size, rois_per_image, 5).zero_()\n # Guard against the case when an image has fewer than max_fg_rois_per_image\n # foreground RoIs\n for i in range(batch_size):\n\n fg_inds = torch.nonzero(max_overlaps[i] >= cfg.TRAIN.FG_THRESH).view(-1)\n fg_num_rois = fg_inds.numel()\n\n # Select background RoIs as those within [BG_THRESH_LO, BG_THRESH_HI)\n bg_inds = torch.nonzero((max_overlaps[i] < cfg.TRAIN.BG_THRESH_HI) &\n (max_overlaps[i] >= cfg.TRAIN.BG_THRESH_LO)).view(-1)\n bg_num_rois = bg_inds.numel()\n\n if fg_num_rois > 0 and bg_num_rois > 0:\n # sampling fg\n fg_rois_per_this_image = min(fg_rois_per_image, fg_num_rois)\n \n # torch.randperm seems has a bug on multi-gpu setting that cause the segfault. \n # See https://github.com/pytorch/pytorch/issues/1868 for more details.\n # use numpy instead.\n #rand_num = torch.randperm(fg_num_rois).long().cuda()\n rand_num = torch.from_numpy(np.random.permutation(fg_num_rois)).type_as(gt_boxes).long()\n fg_inds = fg_inds[rand_num[:fg_rois_per_this_image]]\n\n # sampling bg\n bg_rois_per_this_image = rois_per_image - fg_rois_per_this_image\n\n # Seems torch.rand has a bug, it will generate very large number and make an error. \n # We use numpy rand instead. \n #rand_num = (torch.rand(bg_rois_per_this_image) * bg_num_rois).long().cuda()\n rand_num = np.floor(np.random.rand(bg_rois_per_this_image) * bg_num_rois)\n rand_num = torch.from_numpy(rand_num).type_as(gt_boxes).long()\n bg_inds = bg_inds[rand_num]\n\n elif fg_num_rois > 0 and bg_num_rois == 0:\n # sampling fg\n #rand_num = torch.floor(torch.rand(rois_per_image) * fg_num_rois).long().cuda()\n rand_num = np.floor(np.random.rand(rois_per_image) * fg_num_rois)\n rand_num = torch.from_numpy(rand_num).type_as(gt_boxes).long()\n fg_inds = fg_inds[rand_num]\n fg_rois_per_this_image = rois_per_image\n bg_rois_per_this_image = 0\n elif bg_num_rois > 0 and fg_num_rois == 0:\n # sampling bg\n #rand_num = torch.floor(torch.rand(rois_per_image) * bg_num_rois).long().cuda()\n rand_num = np.floor(np.random.rand(rois_per_image) * bg_num_rois)\n rand_num = torch.from_numpy(rand_num).type_as(gt_boxes).long()\n\n bg_inds = bg_inds[rand_num]\n bg_rois_per_this_image = rois_per_image\n fg_rois_per_this_image = 0\n else:\n raise ValueError(\"bg_num_rois = 0 and fg_num_rois = 0, this should not happen!\")\n \n # The indices that we're selecting (both fg and bg)\n keep_inds = torch.cat([fg_inds, bg_inds], 0)\n\n # Select sampled values from various arrays:\n labels_batch[i].copy_(labels[i][keep_inds])\n\n # Clamp labels for the background RoIs to 0\n if fg_rois_per_this_image < rois_per_image:\n labels_batch[i][fg_rois_per_this_image:] = 0\n\n rois_batch[i] = all_rois[i][keep_inds]\n rois_batch[i,:,0] = i\n\n gt_rois_batch[i] = gt_boxes[i][gt_assignment[i][keep_inds]]\n\n bbox_target_data = self._compute_targets_pytorch(\n rois_batch[:,:,1:5], gt_rois_batch[:,:,:4])\n\n bbox_targets, bbox_inside_weights = \\\n self._get_bbox_regression_labels_pytorch(bbox_target_data, labels_batch, num_classes)\n\n return labels_batch, rois_batch, bbox_targets, bbox_inside_weights\n"
},
{
"path": "lib/model/rpn/proposal_target_layer_cascade_region.py",
"content": "from __future__ import absolute_import\n# --------------------------------------------------------\n# Faster R-CNN\n# Copyright (c) 2015 Microsoft\n# Licensed under The MIT License [see LICENSE for details]\n# Written by Ross Girshick and Sean Bell\n# --------------------------------------------------------\n# --------------------------------------------------------\n# Reorganized and modified by Jianwei Yang and Jiasen Lu\n# --------------------------------------------------------\n\nimport torch\nimport torch.nn as nn\nimport numpy as np\nimport numpy.random as npr\nfrom ..utils.config import cfg\nfrom .bbox_transform import bbox_overlaps_batch, bbox_transform_batch\nimport pdb\n\nclass _ProposalTargetLayer(nn.Module):\n \"\"\"\n Assign object detection proposals to ground-truth targets. Produces proposal\n classification labels and bounding-box regression targets.\n \"\"\"\n\n def __init__(self, nclasses):\n super(_ProposalTargetLayer, self).__init__()\n self._num_classes = nclasses\n self.BBOX_NORMALIZE_MEANS = torch.FloatTensor(cfg.TRAIN.BBOX_NORMALIZE_MEANS)\n self.BBOX_NORMALIZE_STDS = torch.FloatTensor(cfg.TRAIN.BBOX_NORMALIZE_STDS)\n self.BBOX_INSIDE_WEIGHTS = torch.FloatTensor(cfg.TRAIN.BBOX_INSIDE_WEIGHTS)\n\n def forward(self, all_rois, gt_boxes, num_boxes,output_cls_score):\n\n self.BBOX_NORMALIZE_MEANS = self.BBOX_NORMALIZE_MEANS.type_as(gt_boxes)\n self.BBOX_NORMALIZE_STDS = self.BBOX_NORMALIZE_STDS.type_as(gt_boxes)\n self.BBOX_INSIDE_WEIGHTS = self.BBOX_INSIDE_WEIGHTS.type_as(gt_boxes)\n\n gt_boxes_append = gt_boxes.new(gt_boxes.size()).zero_()\n gt_boxes_append[:,:,1:5] = gt_boxes[:,:,:4]\n all_score_append = output_cls_score.new(gt_boxes.size()[0],gt_boxes.size()[1],2).zero_()+1\n\n # Include ground-truth boxes in the set of candidate rois\n all_rois = torch.cat([all_rois, gt_boxes_append], 1)\n all_score = torch.cat([output_cls_score, all_score_append], 1)\n\n num_images = 1\n rois_per_image = int(cfg.TRAIN.BATCH_SIZE / num_images)\n fg_rois_per_image = int(np.round(cfg.TRAIN.FG_FRACTION * rois_per_image))\n fg_rois_per_image = 1 if fg_rois_per_image == 0 else fg_rois_per_image\n\n labels, rois, bbox_targets, bbox_inside_weights, output_bg_score= self._sample_rois_pytorch(\n all_rois, gt_boxes, fg_rois_per_image,\n rois_per_image, self._num_classes, all_score)\n\n bbox_outside_weights = (bbox_inside_weights > 0).float()\n\n return rois, labels, bbox_targets, bbox_inside_weights, bbox_outside_weights, output_bg_score\n\n def backward(self, top, propagate_down, bottom):\n \"\"\"This layer does not propagate gradients.\"\"\"\n pass\n\n def reshape(self, bottom, top):\n \"\"\"Reshaping happens during the call to forward.\"\"\"\n pass\n\n def _get_bbox_regression_labels_pytorch(self, bbox_target_data, labels_batch, num_classes):\n \"\"\"Bounding-box regression targets (bbox_target_data) are stored in a\n compact form b x N x (class, tx, ty, tw, th)\n\n This function expands those targets into the 4-of-4*K representation used\n by the network (i.e. only one class has non-zero targets).\n\n Returns:\n bbox_target (ndarray): b x N x 4K blob of regression targets\n bbox_inside_weights (ndarray): b x N x 4K blob of loss weights\n \"\"\"\n batch_size = labels_batch.size(0)\n rois_per_image = labels_batch.size(1)\n clss = labels_batch\n bbox_targets = bbox_target_data.new(batch_size, rois_per_image, 4).zero_()\n bbox_inside_weights = bbox_target_data.new(bbox_targets.size()).zero_()\n\n for b in range(batch_size):\n # assert clss[b].sum() > 0\n if clss[b].sum() == 0:\n continue\n inds = torch.nonzero(clss[b] > 0).view(-1)\n for i in range(inds.numel()):\n ind = inds[i]\n bbox_targets[b, ind, :] = bbox_target_data[b, ind, :]\n bbox_inside_weights[b, ind, :] = self.BBOX_INSIDE_WEIGHTS\n\n return bbox_targets, bbox_inside_weights\n\n\n def _compute_targets_pytorch(self, ex_rois, gt_rois):\n \"\"\"Compute bounding-box regression targets for an image.\"\"\"\n\n assert ex_rois.size(1) == gt_rois.size(1)\n assert ex_rois.size(2) == 4\n assert gt_rois.size(2) == 4\n\n batch_size = ex_rois.size(0)\n rois_per_image = ex_rois.size(1)\n\n targets = bbox_transform_batch(ex_rois, gt_rois)\n\n if cfg.TRAIN.BBOX_NORMALIZE_TARGETS_PRECOMPUTED:\n # Optionally normalize targets by a precomputed mean and stdev\n targets = ((targets - self.BBOX_NORMALIZE_MEANS.expand_as(targets))\n / self.BBOX_NORMALIZE_STDS.expand_as(targets))\n\n return targets\n\n\n def _sample_rois_pytorch(self, all_rois, gt_boxes, fg_rois_per_image, rois_per_image, num_classes, all_score):\n \"\"\"Generate a random sample of RoIs comprising foreground and background\n examples.\n \"\"\"\n # overlaps: (rois x gt_boxes)\n\n overlaps = bbox_overlaps_batch(all_rois, gt_boxes)\n # max_overlaps = max overlap of (candidate rois with gt_rois)\n max_overlaps, gt_assignment = torch.max(overlaps, 2)\n batch_size = overlaps.size(0)\n num_proposal = overlaps.size(1)\n num_boxes_per_img = overlaps.size(2)\n\n offset = torch.arange(0, batch_size)*gt_boxes.size(1)\n offset = offset.view(-1, 1).type_as(gt_assignment) + gt_assignment\n\n labels = gt_boxes[:,:,4].contiguous().view(-1).index(offset.view(-1))\\\n .view(batch_size, -1)\n\n labels_batch = labels.new(batch_size, rois_per_image).zero_()\n rois_batch = all_rois.new(batch_size, rois_per_image, 5).zero_() # get rois_per_image front of rois\n gt_rois_batch = all_rois.new(batch_size, rois_per_image, 5).zero_()\n output_bg_score = all_score.new(batch_size, rois_per_image, 2).zero_()\n # Guard against the case when an image has fewer than max_fg_rois_per_image\n # foreground RoIs\n for i in range(batch_size):\n\n fg_inds = torch.nonzero(max_overlaps[i] >= cfg.TRAIN.FG_THRESH).view(-1)\n fg_num_rois = fg_inds.numel()\n\n # Select background RoIs as those within [BG_THRESH_LO, BG_THRESH_HI)\n bg_inds = torch.nonzero((max_overlaps[i] < cfg.TRAIN.BG_THRESH_HI) &\n (max_overlaps[i] >= cfg.TRAIN.BG_THRESH_LO)).view(-1)\n bg_num_rois = bg_inds.numel()\n\n if fg_num_rois > 0 and bg_num_rois > 0:\n # sampling fg\n fg_rois_per_this_image = min(fg_rois_per_image, fg_num_rois)\n \n # torch.randperm seems has a bug on multi-gpu setting that cause the segfault. \n # See https://github.com/pytorch/pytorch/issues/1868 for more details.\n # use numpy instead.\n #rand_num = torch.randperm(fg_num_rois).long().cuda()\n rand_num = torch.from_numpy(np.random.permutation(fg_num_rois)).type_as(gt_boxes).long()\n fg_inds = fg_inds[rand_num[:fg_rois_per_this_image]]\n\n # sampling bg\n bg_rois_per_this_image = rois_per_image - fg_rois_per_this_image\n\n # Seems torch.rand has a bug, it will generate very large number and make an error. \n # We use numpy rand instead. \n #rand_num = (torch.rand(bg_rois_per_this_image) * bg_num_rois).long().cuda()\n rand_num = np.floor(np.random.rand(bg_rois_per_this_image) * bg_num_rois)\n rand_num = torch.from_numpy(rand_num).type_as(gt_boxes).long()\n bg_inds = bg_inds[rand_num]\n\n elif fg_num_rois > 0 and bg_num_rois == 0:\n # sampling fg\n #rand_num = torch.floor(torch.rand(rois_per_image) * fg_num_rois).long().cuda()\n rand_num = np.floor(np.random.rand(rois_per_image) * fg_num_rois)\n rand_num = torch.from_numpy(rand_num).type_as(gt_boxes).long()\n fg_inds = fg_inds[rand_num]\n fg_rois_per_this_image = rois_per_image\n bg_rois_per_this_image = 0\n elif bg_num_rois > 0 and fg_num_rois == 0:\n # sampling bg\n #rand_num = torch.floor(torch.rand(rois_per_image) * bg_num_rois).long().cuda()\n rand_num = np.floor(np.random.rand(rois_per_image) * bg_num_rois)\n rand_num = torch.from_numpy(rand_num).type_as(gt_boxes).long()\n\n bg_inds = bg_inds[rand_num]\n bg_rois_per_this_image = rois_per_image\n fg_rois_per_this_image = 0\n else:\n raise ValueError(\"bg_num_rois = 0 and fg_num_rois = 0, this should not happen!\")\n \n # The indices that we're selecting (both fg and bg)\n keep_inds = torch.cat([fg_inds, bg_inds], 0)\n\n # Select sampled values from various arrays:\n labels_batch[i].copy_(labels[i][keep_inds])\n\n # Clamp labels for the background RoIs to 0\n if fg_rois_per_this_image < rois_per_image:\n labels_batch[i][fg_rois_per_this_image:] = 0\n\n rois_batch[i] = all_rois[i][keep_inds]\n rois_batch[i,:,0] = i\n #adding the score\n output_bg_score[i] = all_score[i][keep_inds]\n output_bg_score[i,:,0] = i\n\n gt_rois_batch[i] = gt_boxes[i][gt_assignment[i][keep_inds]]\n\n bbox_target_data = self._compute_targets_pytorch(\n rois_batch[:,:,1:5], gt_rois_batch[:,:,:4])\n\n bbox_targets, bbox_inside_weights = \\\n self._get_bbox_regression_labels_pytorch(bbox_target_data, labels_batch, num_classes)\n\n return labels_batch, rois_batch, bbox_targets, bbox_inside_weights, output_bg_score\n"
},
{
"path": "lib/model/rpn/rpn.py",
"content": "from __future__ import absolute_import\nimport torch\nimport torch.nn as nn\nimport torch.nn.functional as F\nfrom torch.autograd import Variable\n\nfrom model.utils.config import cfg\nfrom .proposal_layer import _ProposalLayer\nfrom .anchor_target_layer import _AnchorTargetLayer\nfrom model.utils.net_utils import _smooth_l1_loss\n\nimport numpy as np\nimport math\nimport pdb\nimport time\n\nclass _RPN(nn.Module):\n \"\"\" region proposal network \"\"\"\n def __init__(self, din):\n super(_RPN, self).__init__()\n \n self.din = din # get depth of input feature map, e.g., 512\n self.anchor_scales = cfg.ANCHOR_SCALES\n self.anchor_ratios = cfg.ANCHOR_RATIOS\n self.feat_stride = cfg.FEAT_STRIDE[0]\n\n # define the convrelu layers processing input feature map\n self.RPN_Conv = nn.Conv2d(self.din, 512, 3, 1, 1, bias=True)\n\n # define bg/fg classifcation score layer\n self.nc_score_out = len(self.anchor_scales) * len(self.anchor_ratios) * 2 # 2(bg/fg) * 9 (anchors)\n self.RPN_cls_score = nn.Conv2d(512, self.nc_score_out, 1, 1, 0)\n\n # define anchor box offset prediction layer\n self.nc_bbox_out = len(self.anchor_scales) * len(self.anchor_ratios) * 4 # 4(coords) * 9 (anchors)\n self.RPN_bbox_pred = nn.Conv2d(512, self.nc_bbox_out, 1, 1, 0)\n\n # define proposal layer\n self.RPN_proposal = _ProposalLayer(self.feat_stride, self.anchor_scales, self.anchor_ratios)\n\n # define anchor target layer\n self.RPN_anchor_target = _AnchorTargetLayer(self.feat_stride, self.anchor_scales, self.anchor_ratios)\n\n self.rpn_loss_cls = 0\n self.rpn_loss_box = 0\n\n @staticmethod\n def reshape(x, d):\n input_shape = x.size()\n x = x.view(\n input_shape[0],\n int(d),\n int(float(input_shape[1] * input_shape[2]) / float(d)),\n input_shape[3]\n )\n return x\n\n def forward(self, base_feat, im_info, gt_boxes, num_boxes):\n\n batch_size = base_feat.size(0)\n\n # return feature map after convrelu layer\n rpn_conv1 = F.relu(self.RPN_Conv(base_feat), inplace=True)\n # get rpn classification score\n rpn_cls_score = self.RPN_cls_score(rpn_conv1)\n\n rpn_cls_score_reshape = self.reshape(rpn_cls_score, 2)\n rpn_cls_prob_reshape = F.softmax(rpn_cls_score_reshape,dim=1)\n rpn_cls_prob = self.reshape(rpn_cls_prob_reshape, self.nc_score_out)\n\n # get rpn offsets to the anchor boxes\n rpn_bbox_pred = self.RPN_bbox_pred(rpn_conv1)\n\n # proposal layer\n cfg_key = 'TRAIN' if self.training else 'TEST'\n\n rois = self.RPN_proposal((rpn_cls_prob.data, rpn_bbox_pred.data,\n im_info, cfg_key))\n\n self.rpn_loss_cls = 0\n self.rpn_loss_box = 0\n\n # generating training labels and build the rpn loss\n if self.training:\n assert gt_boxes is not None\n\n rpn_data = self.RPN_anchor_target((rpn_cls_score.data, gt_boxes, im_info, num_boxes))\n\n # compute classification loss\n rpn_cls_score = rpn_cls_score_reshape.permute(0, 2, 3, 1).contiguous().view(batch_size, -1, 2)\n rpn_label = rpn_data[0].view(batch_size, -1)\n\n rpn_keep = Variable(rpn_label.view(-1).ne(-1).nonzero().view(-1))\n rpn_cls_score = torch.index_select(rpn_cls_score.view(-1,2), 0, rpn_keep)\n rpn_label = torch.index_select(rpn_label.view(-1), 0, rpn_keep.data)\n rpn_label = Variable(rpn_label.long())\n self.rpn_loss_cls = F.cross_entropy(rpn_cls_score, rpn_label)\n fg_cnt = torch.sum(rpn_label.data.ne(0))\n\n rpn_bbox_targets, rpn_bbox_inside_weights, rpn_bbox_outside_weights = rpn_data[1:]\n\n # compute bbox regression loss\n rpn_bbox_inside_weights = Variable(rpn_bbox_inside_weights)\n rpn_bbox_outside_weights = Variable(rpn_bbox_outside_weights)\n rpn_bbox_targets = Variable(rpn_bbox_targets)\n\n self.rpn_loss_box = _smooth_l1_loss(rpn_bbox_pred, rpn_bbox_targets, rpn_bbox_inside_weights,\n rpn_bbox_outside_weights, sigma=3, dim=[1,2,3])\n\n return rois, self.rpn_loss_cls, self.rpn_loss_box\n\n\nclass _RPN_out_pred_label(nn.Module):\n \"\"\" region proposal network \"\"\"\n def __init__(self, din):\n super(_RPN_out_pred_label, self).__init__()\n \n self.din = din # get depth of input feature map, e.g., 512\n self.anchor_scales = cfg.ANCHOR_SCALES\n self.anchor_ratios = cfg.ANCHOR_RATIOS\n self.feat_stride = cfg.FEAT_STRIDE[0]\n\n # define the convrelu layers processing input feature map\n self.RPN_Conv = nn.Conv2d(self.din, 512, 3, 1, 1, bias=True)\n\n # define bg/fg classifcation score layer\n self.nc_score_out = len(self.anchor_scales) * len(self.anchor_ratios) * 2 # 2(bg/fg) * 9 (anchors)\n self.RPN_cls_score = nn.Conv2d(512, self.nc_score_out, 1, 1, 0)\n\n # define anchor box offset prediction layer\n self.nc_bbox_out = len(self.anchor_scales) * len(self.anchor_ratios) * 4 # 4(coords) * 9 (anchors)\n self.RPN_bbox_pred = nn.Conv2d(512, self.nc_bbox_out, 1, 1, 0)\n\n # define proposal layer\n self.RPN_proposal = _ProposalLayer(self.feat_stride, self.anchor_scales, self.anchor_ratios)\n\n # define anchor target layer\n self.RPN_anchor_target = _AnchorTargetLayer(self.feat_stride, self.anchor_scales, self.anchor_ratios)\n\n self.rpn_loss_cls = 0\n self.rpn_loss_box = 0\n\n @staticmethod\n def reshape(x, d):\n input_shape = x.size()\n x = x.view(\n input_shape[0],\n int(d),\n int(float(input_shape[1] * input_shape[2]) / float(d)),\n input_shape[3]\n )\n return x\n\n def forward(self, base_feat, im_info, gt_boxes, num_boxes):\n\n batch_size = base_feat.size(0)\n\n # return feature map after convrelu layer\n rpn_conv1 = F.relu(self.RPN_Conv(base_feat), inplace=True)\n # get rpn classification score\n rpn_cls_score = self.RPN_cls_score(rpn_conv1)\n\n rpn_cls_score_reshape = self.reshape(rpn_cls_score, 2)\n rpn_cls_prob_reshape = F.softmax(rpn_cls_score_reshape)\n rpn_cls_prob = self.reshape(rpn_cls_prob_reshape, self.nc_score_out)\n\n # get rpn offsets to the anchor boxes\n rpn_bbox_pred = self.RPN_bbox_pred(rpn_conv1)\n\n # proposal layer\n cfg_key = 'TRAIN' if self.training else 'TEST'\n\n rois = self.RPN_proposal((rpn_cls_prob.data, rpn_bbox_pred.data,\n im_info, cfg_key))\n\n self.rpn_loss_cls = 0\n self.rpn_loss_box = 0\n\n # generating training labels and build the rpn loss\n if self.training:\n assert gt_boxes is not None\n\n rpn_data = self.RPN_anchor_target((rpn_cls_score.data, gt_boxes, im_info, num_boxes))\n\n # compute classification loss\n rpn_cls_score = rpn_cls_score_reshape.permute(0, 2, 3, 1).contiguous().view(batch_size, -1, 2)\n rpn_label = rpn_data[0].view(batch_size, -1)\n\n rpn_keep = Variable(rpn_label.view(-1).ne(-1).nonzero().view(-1))\n rpn_cls_score = torch.index_select(rpn_cls_score.view(-1,2), 0, rpn_keep)\n rpn_label = torch.index_select(rpn_label.view(-1), 0, rpn_keep.data)\n rpn_label = Variable(rpn_label.long())\n self.rpn_loss_cls = F.cross_entropy(rpn_cls_score, rpn_label)\n fg_cnt = torch.sum(rpn_label.data.ne(0))\n\n rpn_bbox_targets, rpn_bbox_inside_weights, rpn_bbox_outside_weights = rpn_data[1:]\n\n # compute bbox regression loss\n rpn_bbox_inside_weights = Variable(rpn_bbox_inside_weights)\n rpn_bbox_outside_weights = Variable(rpn_bbox_outside_weights)\n rpn_bbox_targets = Variable(rpn_bbox_targets)\n\n self.rpn_loss_box = _smooth_l1_loss(rpn_bbox_pred, rpn_bbox_targets, rpn_bbox_inside_weights,\n rpn_bbox_outside_weights, sigma=3, dim=[1,2,3])\n\n return rois, rpn_cls_prob, self.rpn_loss_cls, self.rpn_loss_box"
},
{
"path": "lib/model/rpn/rpn_region.py",
"content": "from __future__ import absolute_import\nimport torch\nimport torch.nn as nn\nimport torch.nn.functional as F\nfrom torch.autograd import Variable\n\nfrom model.utils.config import cfg\nfrom .proposal_layer_region import _ProposalLayer\nfrom .anchor_target_layer import _AnchorTargetLayer\nfrom model.utils.net_utils import _smooth_l1_loss\n\nimport numpy as np\nimport math\nimport pdb\nimport time\n\nclass _RPN(nn.Module):\n \"\"\" region proposal network \"\"\"\n def __init__(self, din):\n super(_RPN, self).__init__()\n \n self.din = din # get depth of input feature map, e.g., 512\n self.anchor_scales = cfg.ANCHOR_SCALES\n self.anchor_ratios = cfg.ANCHOR_RATIOS\n self.feat_stride = cfg.FEAT_STRIDE[0]\n\n # define the convrelu layers processing input feature map\n self.RPN_Conv = nn.Conv2d(self.din, 512, 3, 1, 1, bias=True)\n\n # define bg/fg classifcation score layer\n self.nc_score_out = len(self.anchor_scales) * len(self.anchor_ratios) * 2 # 2(bg/fg) * 9 (anchors)\n self.RPN_cls_score = nn.Conv2d(512, self.nc_score_out, 1, 1, 0)\n\n # define anchor box offset prediction layer\n self.nc_bbox_out = len(self.anchor_scales) * len(self.anchor_ratios) * 4 # 4(coords) * 9 (anchors)\n self.RPN_bbox_pred = nn.Conv2d(512, self.nc_bbox_out, 1, 1, 0)\n\n # define proposal layer\n self.RPN_proposal = _ProposalLayer(self.feat_stride, self.anchor_scales, self.anchor_ratios)\n\n # define anchor target layer\n self.RPN_anchor_target = _AnchorTargetLayer(self.feat_stride, self.anchor_scales, self.anchor_ratios)\n\n self.rpn_loss_cls = 0\n self.rpn_loss_box = 0\n\n @staticmethod\n def reshape(x, d):\n input_shape = x.size()\n x = x.view(\n input_shape[0],\n int(d),\n int(float(input_shape[1] * input_shape[2]) / float(d)),\n input_shape[3]\n )\n return x\n\n def forward(self, base_feat, im_info, gt_boxes, num_boxes):\n\n batch_size = base_feat.size(0)\n\n # return feature map after convrelu layer\n rpn_conv1 = F.relu(self.RPN_Conv(base_feat), inplace=True)\n # get rpn classification score\n rpn_cls_score = self.RPN_cls_score(rpn_conv1)\n\n rpn_cls_score_reshape = self.reshape(rpn_cls_score, 2)\n rpn_cls_prob_reshape = F.softmax(rpn_cls_score_reshape,dim=1)\n rpn_cls_prob = self.reshape(rpn_cls_prob_reshape, self.nc_score_out)\n\n # get rpn offsets to the anchor boxes\n rpn_bbox_pred = self.RPN_bbox_pred(rpn_conv1)\n\n # proposal layer\n cfg_key = 'TRAIN' if self.training else 'TEST'\n\n rois, output_cls_score= self.RPN_proposal((rpn_cls_prob.data, rpn_bbox_pred.data,\n im_info, cfg_key))\n\n self.rpn_loss_cls = 0\n self.rpn_loss_box = 0\n\n # generating training labels and build the rpn loss\n if self.training:\n assert gt_boxes is not None\n\n rpn_data = self.RPN_anchor_target((rpn_cls_score.data, gt_boxes, im_info, num_boxes))\n\n # compute classification loss\n rpn_cls_score = rpn_cls_score_reshape.permute(0, 2, 3, 1).contiguous().view(batch_size, -1, 2)\n rpn_label = rpn_data[0].view(batch_size, -1)\n\n rpn_keep = Variable(rpn_label.view(-1).ne(-1).nonzero().view(-1))\n rpn_cls_score = torch.index_select(rpn_cls_score.view(-1,2), 0, rpn_keep)\n rpn_label = torch.index_select(rpn_label.view(-1), 0, rpn_keep.data)\n rpn_label = Variable(rpn_label.long())\n self.rpn_loss_cls = F.cross_entropy(rpn_cls_score, rpn_label)\n fg_cnt = torch.sum(rpn_label.data.ne(0))\n\n rpn_bbox_targets, rpn_bbox_inside_weights, rpn_bbox_outside_weights = rpn_data[1:]\n\n # compute bbox regression loss\n rpn_bbox_inside_weights = Variable(rpn_bbox_inside_weights)\n rpn_bbox_outside_weights = Variable(rpn_bbox_outside_weights)\n rpn_bbox_targets = Variable(rpn_bbox_targets)\n\n self.rpn_loss_box = _smooth_l1_loss(rpn_bbox_pred, rpn_bbox_targets, rpn_bbox_inside_weights,\n rpn_bbox_outside_weights, sigma=3, dim=[1,2,3])\n\n return rois, output_cls_score, self.rpn_loss_cls, self.rpn_loss_box\n\n\n"
},
{
"path": "lib/model/utils/.gitignore",
"content": "*.c\n*.cpp\n*.so\n"
},
{
"path": "lib/model/utils/__init__.py",
"content": ""
},
{
"path": "lib/model/utils/bbox.pyx",
"content": "# --------------------------------------------------------\n# Fast R-CNN\n# Copyright (c) 2015 Microsoft\n# Licensed under The MIT License [see LICENSE for details]\n# Written by Sergey Karayev\n# --------------------------------------------------------\n\ncimport cython\nimport numpy as np\ncimport numpy as np\n\nDTYPE = np.float\nctypedef np.float_t DTYPE_t\n\ndef bbox_overlaps(np.ndarray[DTYPE_t, ndim=2] boxes,\n np.ndarray[DTYPE_t, ndim=2] query_boxes):\n return bbox_overlaps_c(boxes, query_boxes)\n\ncdef np.ndarray[DTYPE_t, ndim=2] bbox_overlaps_c(\n np.ndarray[DTYPE_t, ndim=2] boxes,\n np.ndarray[DTYPE_t, ndim=2] query_boxes):\n \"\"\"\n Parameters\n ----------\n boxes: (N, 4) ndarray of float\n query_boxes: (K, 4) ndarray of float\n Returns\n -------\n overlaps: (N, K) ndarray of overlap between boxes and query_boxes\n \"\"\"\n cdef unsigned int N = boxes.shape[0]\n cdef unsigned int K = query_boxes.shape[0]\n cdef np.ndarray[DTYPE_t, ndim=2] overlaps = np.zeros((N, K), dtype=DTYPE)\n cdef DTYPE_t iw, ih, box_area\n cdef DTYPE_t ua\n cdef unsigned int k, n\n for k in range(K):\n box_area = (\n (query_boxes[k, 2] - query_boxes[k, 0] + 1) *\n (query_boxes[k, 3] - query_boxes[k, 1] + 1)\n )\n for n in range(N):\n iw = (\n min(boxes[n, 2], query_boxes[k, 2]) -\n max(boxes[n, 0], query_boxes[k, 0]) + 1\n )\n if iw > 0:\n ih = (\n min(boxes[n, 3], query_boxes[k, 3]) -\n max(boxes[n, 1], query_boxes[k, 1]) + 1\n )\n if ih > 0:\n ua = float(\n (boxes[n, 2] - boxes[n, 0] + 1) *\n (boxes[n, 3] - boxes[n, 1] + 1) +\n box_area - iw * ih\n )\n overlaps[n, k] = iw * ih / ua\n return overlaps\n\n\ndef bbox_intersections(\n np.ndarray[DTYPE_t, ndim=2] boxes,\n np.ndarray[DTYPE_t, ndim=2] query_boxes):\n return bbox_intersections_c(boxes, query_boxes)\n\n\ncdef np.ndarray[DTYPE_t, ndim=2] bbox_intersections_c(\n np.ndarray[DTYPE_t, ndim=2] boxes,\n np.ndarray[DTYPE_t, ndim=2] query_boxes):\n \"\"\"\n For each query box compute the intersection ratio covered by boxes\n ----------\n Parameters\n ----------\n boxes: (N, 4) ndarray of float\n query_boxes: (K, 4) ndarray of float\n Returns\n -------\n overlaps: (N, K) ndarray of intersec between boxes and query_boxes\n \"\"\"\n cdef unsigned int N = boxes.shape[0]\n cdef unsigned int K = query_boxes.shape[0]\n cdef np.ndarray[DTYPE_t, ndim=2] intersec = np.zeros((N, K), dtype=DTYPE)\n cdef DTYPE_t iw, ih, box_area\n cdef DTYPE_t ua\n cdef unsigned int k, n\n for k in range(K):\n box_area = (\n (query_boxes[k, 2] - query_boxes[k, 0] + 1) *\n (query_boxes[k, 3] - query_boxes[k, 1] + 1)\n )\n for n in range(N):\n iw = (\n min(boxes[n, 2], query_boxes[k, 2]) -\n max(boxes[n, 0], query_boxes[k, 0]) + 1\n )\n if iw > 0:\n ih = (\n min(boxes[n, 3], query_boxes[k, 3]) -\n max(boxes[n, 1], query_boxes[k, 1]) + 1\n )\n if ih > 0:\n intersec[n, k] = iw * ih / box_area\n return intersec"
},
{
"path": "lib/model/utils/blob.py",
"content": "# --------------------------------------------------------\n# Fast R-CNN\n# Copyright (c) 2015 Microsoft\n# Licensed under The MIT License [see LICENSE for details]\n# Written by Ross Girshick\n# --------------------------------------------------------\n\n\"\"\"Blob helper functions.\"\"\"\n\nimport numpy as np\n# from scipy.misc import imread, imresize\nimport cv2\n\ntry:\n xrange # Python 2\nexcept NameError:\n xrange = range # Python 3\n\n\ndef im_list_to_blob(ims):\n \"\"\"Convert a list of images into a network input.\n\n Assumes images are already prepared (means subtracted, BGR order, ...).\n \"\"\"\n max_shape = np.array([im.shape for im in ims]).max(axis=0)\n num_images = len(ims)\n blob = np.zeros((num_images, max_shape[0], max_shape[1], 3),\n dtype=np.float32)\n for i in xrange(num_images):\n im = ims[i]\n blob[i, 0:im.shape[0], 0:im.shape[1], :] = im\n\n return blob\n\ndef prep_im_for_blob(im, pixel_means, target_size, max_size):\n \"\"\"Mean subtract and scale an image for use in a blob.\"\"\"\n\n im = im.astype(np.float32, copy=False)\n im -= pixel_means\n # im = im[:, :, ::-1]\n im_shape = im.shape\n im_size_min = np.min(im_shape[0:2])\n im_size_max = np.max(im_shape[0:2])\n im_scale = float(target_size) / float(im_size_min)\n # Prevent the biggest axis from being more than MAX_SIZE\n # if np.round(im_scale * im_size_max) > max_size:\n # im_scale = float(max_size) / float(im_size_max)\n # im = imresize(im, im_scale)\n im = cv2.resize(im, None, None, fx=im_scale, fy=im_scale,\n interpolation=cv2.INTER_LINEAR)\n\n return im, im_scale\n"
},
{
"path": "lib/model/utils/config.py",
"content": "from __future__ import absolute_import\nfrom __future__ import division\nfrom __future__ import print_function\n\nimport os\nimport os.path as osp\nimport numpy as np\n# `pip install easydict` if you don't have it\nfrom easydict import EasyDict as edict\n\n__C = edict()\n# Consumers can get config by:\n# from fast_rcnn_config import cfg\ncfg = __C\n\n#\n# Training options\n#\n__C.TRAIN = edict()\n\n# Initial learning rate\n__C.TRAIN.LEARNING_RATE = 0.001\n\n__C.TRAIN.META_TYPE = 1\n\n##### set classes ###\n__C.TRAIN.ALLCLASSES_FIRST = ['aeroplane', 'bicycle', 'boat', 'bottle', 'car', 'cat', 'chair', 'diningtable', 'dog',\n 'horse', 'person', 'pottedplant', 'sheep', 'train', 'tvmonitor', 'bird', 'bus', 'cow',\n 'motorbike', 'sofa']\n\n__C.TRAIN.BASECLASSES_FIRST = ['aeroplane', 'bicycle', 'boat',\n 'bottle', 'car', 'cat', 'chair',\n 'diningtable', 'dog', 'horse',\n 'person', 'pottedplant',\n 'sheep', 'train', 'tvmonitor']\n\n__C.TRAIN.ALLCLASSES_SECOND = ['bicycle', 'bird', 'boat', 'bus', 'car', 'cat', 'chair', 'diningtable', 'dog', 'motorbike',\n 'person', 'pottedplant', 'sheep', 'train', 'tvmonitor', 'aeroplane', 'bottle', 'cow',\n 'horse', 'sofa']\n\n__C.TRAIN.BASECLASSES_SECOND = ['bicycle', 'bird', 'boat', 'bus',\n 'car','cat', 'chair','diningtable',\n 'dog', 'motorbike','person', 'pottedplant',\n 'sheep','train', 'tvmonitor']\n\n__C.TRAIN.ALLCLASSES_THIRD = ['aeroplane', 'bicycle', 'bird', 'bottle', 'bus', 'car', 'chair', 'cow', 'diningtable',\n 'dog', 'horse', 'person', 'pottedplant', 'train', 'tvmonitor', 'boat', 'cat', 'motorbike',\n 'sheep', 'sofa']\n\n__C.TRAIN.BASECLASSES_THIRD = ['aeroplane', 'bicycle', 'bird','bottle',\n 'bus', 'car', 'chair','cow', 'diningtable',\n 'dog', 'horse','person', 'pottedplant', 'train', 'tvmonitor']\n####\n# Momentum\n__C.TRAIN.MOMENTUM = 0.9\n\n# Weight decay, for regularization\n__C.TRAIN.WEIGHT_DECAY = 0.0005\n\n# Factor for reducing the learning rate\n__C.TRAIN.GAMMA = 0.1\n\n# Step size for reducing the learning rate, currently only support one step\n__C.TRAIN.STEPSIZE = [30000]\n\n# Iteration intervals for showing the loss during training, on command line interface\n__C.TRAIN.DISPLAY = 10\n\n# Whether to double the learning rate for bias\n__C.TRAIN.DOUBLE_BIAS = True\n\n# Whether to initialize the weights with truncated normal distribution\n__C.TRAIN.TRUNCATED = False\n\n# Whether to have weight decay on bias as well\n__C.TRAIN.BIAS_DECAY = False\n\n# Whether to add ground truth boxes to the pool when sampling regions\n__C.TRAIN.USE_GT = False\n\n# Whether to use aspect-ratio grouping of training images, introduced merely for saving\n# GPU memory\n__C.TRAIN.ASPECT_GROUPING = False\n\n# The number of snapshots kept, older ones are deleted to save space\n__C.TRAIN.SNAPSHOT_KEPT = 3\n\n# The time interval for saving tensorflow summaries\n__C.TRAIN.SUMMARY_INTERVAL = 180\n\n# Scale to use during training (can list multiple scales)\n# The scale is the pixel size of an image's shortest side\n__C.TRAIN.SCALES = (600,)\n\n# Max pixel size of the longest side of a scaled input image\n__C.TRAIN.MAX_SIZE = 1000\n\n# Trim size for input images to create minibatch\n__C.TRAIN.TRIM_HEIGHT = 600\n__C.TRAIN.TRIM_WIDTH = 600\n\n# Images to use per minibatch\n__C.TRAIN.IMS_PER_BATCH = 1\n\n# Minibatch size (number of regions of interest [ROIs])\n__C.TRAIN.BATCH_SIZE = 128\n\n# Fraction of minibatch that is labeled foreground (i.e. class > 0)\n__C.TRAIN.FG_FRACTION = 0.25\n\n# Overlap threshold for a ROI to be considered foreground (if >= FG_THRESH)\n__C.TRAIN.FG_THRESH = 0.5\n\n# Overlap threshold for a ROI to be considered background (class = 0 if\n# overlap in [LO, HI))\n__C.TRAIN.BG_THRESH_HI = 0.5\n__C.TRAIN.BG_THRESH_LO = 0.1\n\n# Use horizontally-flipped images during training?\n__C.TRAIN.USE_FLIPPED = True\n\n# Train bounding-box regressors\n__C.TRAIN.BBOX_REG = True\n__C.TRAIN.RCNN_BBOX_WEIGHT = 1\n\n# Overlap required between a ROI and ground-truth box in order for that ROI to\n# be used as a bounding-box regression training example\n__C.TRAIN.BBOX_THRESH = 0.5\n\n# Iterations between snapshots\n__C.TRAIN.SNAPSHOT_ITERS = 5000\n\n# solver.prototxt specifies the snapshot path prefix, this adds an optional\n# infix to yield the path: [_]_iters_XYZ.caffemodel\n__C.TRAIN.SNAPSHOT_PREFIX = 'res101_faster_rcnn'\n# __C.TRAIN.SNAPSHOT_INFIX = ''\n\n# Use a prefetch thread in roi_data_layer.layer\n# So far I haven't found this useful; likely more engineering work is required\n# __C.TRAIN.USE_PREFETCH = False\n\n# Normalize the targets (subtract empirical mean, divide by empirical stddev)\n__C.TRAIN.BBOX_NORMALIZE_TARGETS = True\n# Deprecated (inside weights)\n__C.TRAIN.BBOX_INSIDE_WEIGHTS = (1.0, 1.0, 1.0, 1.0)\n# Normalize the targets using \"precomputed\" (or made up) means and stdevs\n# (BBOX_NORMALIZE_TARGETS must also be True)\n__C.TRAIN.BBOX_NORMALIZE_TARGETS_PRECOMPUTED = True\n__C.TRAIN.BBOX_NORMALIZE_MEANS = (0.0, 0.0, 0.0, 0.0)\n__C.TRAIN.BBOX_NORMALIZE_STDS = (0.1, 0.1, 0.2, 0.2)\n\n# Train using these proposals\n__C.TRAIN.PROPOSAL_METHOD = 'gt'\n\n# Make minibatches from images that have similar aspect ratios (i.e. both\n# tall and thin or both short and wide) in order to avoid wasting computation\n# on zero-padding.\n\n# Use RPN to detect objects\n__C.TRAIN.HAS_RPN = True\n# IOU >= thresh: positive example\n__C.TRAIN.RPN_POSITIVE_OVERLAP = 0.7\n# IOU < thresh: negative example\n__C.TRAIN.RPN_NEGATIVE_OVERLAP = 0.3\n# If an anchor statisfied by positive and negative conditions set to negative\n__C.TRAIN.RPN_CLOBBER_POSITIVES = False\n# Max number of foreground examples\n__C.TRAIN.RPN_FG_FRACTION = 0.5\n# Total number of examples\n__C.TRAIN.RPN_BATCHSIZE = 256\n# NMS threshold used on RPN proposals\n__C.TRAIN.RPN_NMS_THRESH = 0.7\n# Number of top scoring boxes to keep before apply NMS to RPN proposals\n__C.TRAIN.RPN_PRE_NMS_TOP_N = 12000\n# Number of top scoring boxes to keep after applying NMS to RPN proposals\n__C.TRAIN.RPN_POST_NMS_TOP_N = 2000\n# Proposal height and width both need to be greater than RPN_MIN_SIZE (at orig image scale)\n__C.TRAIN.RPN_MIN_SIZE = 8\n# Deprecated (outside weights)\n__C.TRAIN.RPN_BBOX_INSIDE_WEIGHTS = (1.0, 1.0, 1.0, 1.0)\n# Give the positive RPN examples weight of p * 1 / {num positives}\n# and give negatives a weight of (1 - p)\n# Set to -1.0 to use uniform example weighting\n__C.TRAIN.RPN_POSITIVE_WEIGHT = -1.0\n# Whether to use all ground truth bounding boxes for training,\n# For COCO, setting USE_ALL_GT to False will exclude boxes that are flagged as ''iscrowd''\n__C.TRAIN.USE_ALL_GT = True\n\n# Whether to tune the batch normalization parameters during training\n__C.TRAIN.BN_TRAIN = False\n\n#\n# Testing options\n#\n__C.TEST = edict()\n\n# Scale to use during testing (can NOT list multiple scales)\n# The scale is the pixel size of an image's shortest side\n__C.TEST.SCALES = (600,)\n\n# Max pixel size of the longest side of a scaled input image\n__C.TEST.MAX_SIZE = 1000\n\n# Overlap threshold used for non-maximum suppression (suppress boxes with\n# IoU >= this threshold)\n__C.TEST.NMS = 0.3\n\n# Experimental: treat the (K+1) units in the cls_score layer as linear\n# predictors (trained, eg, with one-vs-rest SVMs).\n__C.TEST.SVM = False\n\n# Test using bounding-box regressors\n__C.TEST.BBOX_REG = True\n\n# Propose boxes\n__C.TEST.HAS_RPN = False\n\n# Test using these proposals\n__C.TEST.PROPOSAL_METHOD = 'gt'\n\n## NMS threshold used on RPN proposals\n__C.TEST.RPN_NMS_THRESH = 0.7\n## Number of top scoring boxes to keep before apply NMS to RPN proposals\n__C.TEST.RPN_PRE_NMS_TOP_N = 6000\n\n## Number of top scoring boxes to keep after applying NMS to RPN proposals\n__C.TEST.RPN_POST_NMS_TOP_N = 300\n\n# Proposal height and width both need to be greater than RPN_MIN_SIZE (at orig image scale)\n__C.TEST.RPN_MIN_SIZE = 16\n\n# Testing mode, default to be 'nms', 'top' is slower but better\n# See report for details\n__C.TEST.MODE = 'nms'\n\n# Only useful when TEST.MODE is 'top', specifies the number of top proposals to select\n__C.TEST.RPN_TOP_N = 5000\n\n#\n# ResNet options\n#\n\n__C.RESNET = edict()\n\n# Option to set if max-pooling is appended after crop_and_resize.\n# if true, the region will be resized to a square of 2xPOOLING_SIZE,\n# then 2x2 max-pooling is applied; otherwise the region will be directly\n# resized to a square of POOLING_SIZE\n__C.RESNET.MAX_POOL = False\n\n# Number of fixed blocks during training, by default the FIRST of all 4 blocks is fixed\n# Range: 0 (none) to 4 (all)\n__C.RESNET.FIXED_BLOCKS = 2\n\n#\n# MobileNet options\n#\n\n__C.MOBILENET = edict()\n\n# Whether to regularize the depth-wise filters during training\n__C.MOBILENET.REGU_DEPTH = False\n\n# Number of fixed layers during training, by default the FIRST of all 14 layers is fixed\n# Range: 0 (none) to 12 (all)\n__C.MOBILENET.FIXED_LAYERS = 5\n\n# Weight decay for the mobilenet weights\n__C.MOBILENET.WEIGHT_DECAY = 0.00004\n\n# Depth multiplier\n__C.MOBILENET.DEPTH_MULTIPLIER = 1.\n\n#\n# MISC\n#\n\n# The mapping from image coordinates to feature map coordinates might cause\n# some boxes that are distinct in image space to become identical in feature\n# coordinates. If DEDUP_BOXES > 0, then DEDUP_BOXES is used as the scale factor\n# for identifying duplicate boxes.\n# 1/16 is correct for {Alex,Caffe}Net, VGG_CNN_M_1024, and VGG16\n__C.DEDUP_BOXES = 1. / 16.\n\n# Pixel mean values (BGR order) as a (1, 1, 3) array\n# We use the same pixel mean for all networks even though it's not exactly what\n# they were trained with\n__C.PIXEL_MEANS = np.array([[[102.9801, 115.9465, 122.7717]]])\n\n# For reproducibility\n__C.RNG_SEED = 3\n\n# A small number that's used many times\n__C.EPS = 1e-14\n\n# Root directory of project\n__C.ROOT_DIR = osp.abspath(osp.join(osp.dirname(__file__), '..', '..', '..'))\n__C.ROOT_DATA = '/'\n# Data directory\n#__C.DATA_DIR = osp.abspath(osp.join(__C.ROOT_DIR, 'data'))\n__C.DATA_DIR = './data'\n\n# Name (or path to) the matlab executable\n__C.MATLAB = 'matlab'\n\n# Place outputs under an experiments directory\n__C.EXP_DIR = 'default'\n\n# Use GPU implementation of non-maximum suppression\n__C.USE_GPU_NMS = True\n\n# Default GPU device id\n__C.GPU_ID = 0\n\n__C.POOLING_MODE = 'align'\n\n# Size of the pooled region after RoI pooling\n__C.POOLING_SIZE = 7\n\n# Maximal number of gt rois in an image during Training\n__C.MAX_NUM_GT_BOXES = 20\n\n# Anchor scales for RPN\n__C.ANCHOR_SCALES = [2,4,8,16,32]#[8,16,32]\n\n# Anchor ratios for RPN\n__C.ANCHOR_RATIOS = [0.5,1,2]\n\n# Feature stride for RPN\n__C.FEAT_STRIDE = [16, ]\n\n__C.CUDA = False\n\n__C.CROP_RESIZE_WITH_MAX_POOL = True\n\nimport pdb\ndef get_output_dir(imdb, weights_filename):\n \"\"\"Return the directory where experimental artifacts are placed.\n If the directory does not exist, it is created.\n\n A canonical path is built using the name from an imdb and a network\n (if not None).\n \"\"\"\n outdir = osp.abspath(osp.join(__C.ROOT_DIR, 'output', __C.EXP_DIR, imdb.name))\n if weights_filename is None:\n weights_filename = 'default'\n outdir = osp.join(outdir, weights_filename)\n if not os.path.exists(outdir):\n os.makedirs(outdir)\n return outdir\n\n\ndef get_output_tb_dir(imdb, weights_filename):\n \"\"\"Return the directory where tensorflow summaries are placed.\n If the directory does not exist, it is created.\n\n A canonical path is built using the name from an imdb and a network\n (if not None).\n \"\"\"\n outdir = osp.abspath(osp.join(__C.ROOT_DIR, 'tensorboard', __C.EXP_DIR, imdb.name))\n if weights_filename is None:\n weights_filename = 'default'\n outdir = osp.join(outdir, weights_filename)\n if not os.path.exists(outdir):\n os.makedirs(outdir)\n return outdir\n\n\ndef _merge_a_into_b(a, b):\n \"\"\"Merge config dictionary a into config dictionary b, clobbering the\n options in b whenever they are also specified in a.\n \"\"\"\n if type(a) is not edict:\n return\n\n for k, v in a.items():\n # a must specify keys that are in b\n if k not in b:\n raise KeyError('{} is not a valid config key'.format(k))\n\n # the types must match, too\n old_type = type(b[k])\n if old_type is not type(v):\n if isinstance(b[k], np.ndarray):\n v = np.array(v, dtype=b[k].dtype)\n else:\n raise ValueError(('Type mismatch ({} vs. {}) '\n 'for config key: {}').format(type(b[k]),\n type(v), k))\n\n # recursively merge dicts\n if type(v) is edict:\n try:\n _merge_a_into_b(a[k], b[k])\n except:\n print(('Error under config key: {}'.format(k)))\n raise\n else:\n b[k] = v\n\n\ndef cfg_from_file(filename):\n \"\"\"Load a config file and merge it into the default options.\"\"\"\n import yaml\n with open(filename, 'r') as f:\n yaml_cfg = edict(yaml.load(f))\n\n _merge_a_into_b(yaml_cfg, __C)\n\n\ndef cfg_from_list(cfg_list):\n \"\"\"Set config keys via list (e.g., from command line).\"\"\"\n from ast import literal_eval\n assert len(cfg_list) % 2 == 0\n for k, v in zip(cfg_list[0::2], cfg_list[1::2]):\n key_list = k.split('.')\n d = __C\n for subkey in key_list[:-1]:\n assert subkey in d\n d = d[subkey]\n subkey = key_list[-1]\n assert subkey in d\n try:\n value = literal_eval(v)\n except:\n # handle the case when v is a string literal\n value = v\n assert type(value) == type(d[subkey]), \\\n 'type {} does not match original type {}'.format(\n type(value), type(d[subkey]))\n d[subkey] = value\n"
},
{
"path": "lib/model/utils/net_utils.py",
"content": "import torch\nimport torch.nn as nn\nimport torch.nn.functional as F\nfrom torch.autograd import Variable\nimport numpy as np\nimport torchvision.models as models\nfrom model.utils.config import cfg\nfrom model.roi_crop.functions.roi_crop import RoICropFunction\nimport cv2\nimport pdb\nimport random\n\ndef save_net(fname, net):\n import h5py\n h5f = h5py.File(fname, mode='w')\n for k, v in net.state_dict().items():\n h5f.create_dataset(k, data=v.cpu().numpy())\n\ndef load_net(fname, net):\n import h5py\n h5f = h5py.File(fname, mode='r')\n for k, v in net.state_dict().items():\n param = torch.from_numpy(np.asarray(h5f[k]))\n v.copy_(param)\n\ndef weights_normal_init(model, dev=0.01):\n if isinstance(model, list):\n for m in model:\n weights_normal_init(m, dev)\n else:\n for m in model.modules():\n if isinstance(m, nn.Conv2d):\n m.weight.data.normal_(0.0, dev)\n elif isinstance(m, nn.Linear):\n m.weight.data.normal_(0.0, dev)\n\n\ndef clip_gradient(model, clip_norm):\n \"\"\"Computes a gradient clipping coefficient based on gradient norm.\"\"\"\n totalnorm = 0\n for p in model.parameters():\n if p.requires_grad:\n modulenorm = p.grad.data.norm()\n totalnorm += modulenorm ** 2\n totalnorm = np.sqrt(totalnorm)\n\n norm = clip_norm / max(totalnorm, clip_norm)\n for p in model.parameters():\n if p.requires_grad:\n p.grad.mul_(norm)\n\ndef vis_detections(im, class_name, dets, thresh=0.8):\n \"\"\"Visual debugging of detections.\"\"\"\n class_name = class_name.split('.')[0]\n for i in range(np.minimum(10, dets.shape[0])):\n bbox = tuple(int(np.round(x)) for x in dets[i, :4])\n score = dets[i, -1]\n if score > thresh:\n cv2.rectangle(im, bbox[0:2], bbox[2:4], (255, 255, 0), 1)\n text_size = cv2.getTextSize('%s: %.3f' % (class_name, score), cv2.FONT_HERSHEY_PLAIN, 0.8, 1)\n point = (bbox[0] + text_size[0][0], bbox[1] + text_size[0][1] + text_size[1])\n cv2.rectangle(im, bbox[0:2], point, (255, 255, 0), -1)\n cv2.putText(im, '%s: %.3f' % (class_name, score), (bbox[0], bbox[1] + 10), cv2.FONT_HERSHEY_PLAIN,\n 0.8, (0, 0, 255), thickness=1)\n # cv2.putText(im, '%s' % (class_name), (bbox[0], bbox[1] + 10), cv2.FONT_HERSHEY_PLAIN,\n # 0.8, (0, 0, 255), thickness=1)\n return im\n\ndef vis_detections_label_only(im, class_name, dets, thresh=0.8):\n \"\"\"Visual debugging of detections.\"\"\"\n class_name = class_name.split('.')[0]\n for i in range(np.minimum(10, dets.shape[0])):\n bbox = tuple(int(np.round(x)) for x in dets[i, :4])\n score = dets[i, -1]\n if score > thresh:\n cv2.rectangle(im, bbox[0:2], bbox[2:4], (0, 255, 0), 2)\n text_size = cv2.getTextSize('%s' % (class_name), cv2.FONT_HERSHEY_COMPLEX, 0.5, 1)\n point = (bbox[0] + text_size[0][0], bbox[1] + text_size[0][1] + text_size[1])\n cv2.rectangle(im, bbox[0:2], point, (0, 255, 0), -1)\n cv2.putText(im, '%s' % (class_name), (bbox[0], bbox[1] + 10), cv2.FONT_HERSHEY_COMPLEX,\n 0.5, (0, 0, 0), thickness=1)\n # cv2.putText(im, '%s' % (class_name), (bbox[0], bbox[1] + 10), cv2.FONT_HERSHEY_PLAIN,\n # 0.8, (0, 0, 255), thickness=1)\n return im\n\n\n\ndef adjust_learning_rate(optimizer, decay=0.1):\n \"\"\"Sets the learning rate to the initial LR decayed by 0.5 every 20 epochs\"\"\"\n for param_group in optimizer.param_groups:\n param_group['lr'] = decay * param_group['lr']\n\n\ndef save_checkpoint(state, filename):\n torch.save(state, filename)\n\ndef _smooth_l1_loss(bbox_pred, bbox_targets, bbox_inside_weights, bbox_outside_weights, sigma=1.0, dim=[1]):\n \n sigma_2 = sigma ** 2\n box_diff = bbox_pred - bbox_targets\n in_box_diff = bbox_inside_weights * box_diff\n abs_in_box_diff = torch.abs(in_box_diff)\n smoothL1_sign = (abs_in_box_diff < 1. / sigma_2).detach().float()\n in_loss_box = torch.pow(in_box_diff, 2) * (sigma_2 / 2.) * smoothL1_sign \\\n + (abs_in_box_diff - (0.5 / sigma_2)) * (1. - smoothL1_sign)\n out_loss_box = bbox_outside_weights * in_loss_box\n loss_box = out_loss_box\n for i in sorted(dim, reverse=True):\n loss_box = loss_box.sum(i)\n loss_box = loss_box.mean()\n return loss_box\n\ndef _crop_pool_layer(bottom, rois, max_pool=True):\n # code modified from \n # https://github.com/ruotianluo/pytorch-faster-rcnn\n # implement it using stn\n # box to affine\n # input (x1,y1,x2,y2)\n \"\"\"\n [ x2-x1 x1 + x2 - W + 1 ]\n [ ----- 0 --------------- ]\n [ W - 1 W - 1 ]\n [ ]\n [ y2-y1 y1 + y2 - H + 1 ]\n [ 0 ----- --------------- ]\n [ H - 1 H - 1 ]\n \"\"\"\n rois = rois.detach()\n batch_size = bottom.size(0)\n D = bottom.size(1)\n H = bottom.size(2)\n W = bottom.size(3)\n roi_per_batch = rois.size(0) / batch_size\n x1 = rois[:, 1::4] / 16.0\n y1 = rois[:, 2::4] / 16.0\n x2 = rois[:, 3::4] / 16.0\n y2 = rois[:, 4::4] / 16.0\n\n height = bottom.size(2)\n width = bottom.size(3)\n\n # affine theta\n zero = Variable(rois.data.new(rois.size(0), 1).zero_())\n theta = torch.cat([\\\n (x2 - x1) / (width - 1),\n zero,\n (x1 + x2 - width + 1) / (width - 1),\n zero,\n (y2 - y1) / (height - 1),\n (y1 + y2 - height + 1) / (height - 1)], 1).view(-1, 2, 3)\n\n if max_pool:\n pre_pool_size = cfg.POOLING_SIZE * 2\n grid = F.affine_grid(theta, torch.Size((rois.size(0), 1, pre_pool_size, pre_pool_size)))\n bottom = bottom.view(1, batch_size, D, H, W).contiguous().expand(roi_per_batch, batch_size, D, H, W)\\\n .contiguous().view(-1, D, H, W)\n crops = F.grid_sample(bottom, grid)\n crops = F.max_pool2d(crops, 2, 2)\n else:\n grid = F.affine_grid(theta, torch.Size((rois.size(0), 1, cfg.POOLING_SIZE, cfg.POOLING_SIZE)))\n bottom = bottom.view(1, batch_size, D, H, W).contiguous().expand(roi_per_batch, batch_size, D, H, W)\\\n .contiguous().view(-1, D, H, W)\n crops = F.grid_sample(bottom, grid)\n \n return crops, grid\n\ndef _affine_grid_gen(rois, input_size, grid_size):\n\n rois = rois.detach()\n x1 = rois[:, 1::4] / 16.0\n y1 = rois[:, 2::4] / 16.0\n x2 = rois[:, 3::4] / 16.0\n y2 = rois[:, 4::4] / 16.0\n\n height = input_size[0]\n width = input_size[1]\n\n zero = Variable(rois.data.new(rois.size(0), 1).zero_())\n theta = torch.cat([\\\n (x2 - x1) / (width - 1),\n zero,\n (x1 + x2 - width + 1) / (width - 1),\n zero,\n (y2 - y1) / (height - 1),\n (y1 + y2 - height + 1) / (height - 1)], 1).view(-1, 2, 3)\n\n grid = F.affine_grid(theta, torch.Size((rois.size(0), 1, grid_size, grid_size)))\n\n return grid\n\ndef _affine_theta(rois, input_size):\n\n rois = rois.detach()\n x1 = rois[:, 1::4] / 16.0\n y1 = rois[:, 2::4] / 16.0\n x2 = rois[:, 3::4] / 16.0\n y2 = rois[:, 4::4] / 16.0\n\n height = input_size[0]\n width = input_size[1]\n\n zero = Variable(rois.data.new(rois.size(0), 1).zero_())\n\n # theta = torch.cat([\\\n # (x2 - x1) / (width - 1),\n # zero,\n # (x1 + x2 - width + 1) / (width - 1),\n # zero,\n # (y2 - y1) / (height - 1),\n # (y1 + y2 - height + 1) / (height - 1)], 1).view(-1, 2, 3)\n\n theta = torch.cat([\\\n (y2 - y1) / (height - 1),\n zero,\n (y1 + y2 - height + 1) / (height - 1),\n zero,\n (x2 - x1) / (width - 1),\n (x1 + x2 - width + 1) / (width - 1)], 1).view(-1, 2, 3)\n\n return theta\n\ndef compare_grid_sample():\n # do gradcheck\n N = random.randint(1, 8)\n C = 2 # random.randint(1, 8)\n H = 5 # random.randint(1, 8)\n W = 4 # random.randint(1, 8)\n input = Variable(torch.randn(N, C, H, W).cuda(), requires_grad=True)\n input_p = input.clone().data.contiguous()\n \n grid = Variable(torch.randn(N, H, W, 2).cuda(), requires_grad=True)\n grid_clone = grid.clone().contiguous()\n\n out_offcial = F.grid_sample(input, grid) \n grad_outputs = Variable(torch.rand(out_offcial.size()).cuda())\n grad_outputs_clone = grad_outputs.clone().contiguous()\n grad_inputs = torch.autograd.grad(out_offcial, (input, grid), grad_outputs.contiguous())\n grad_input_off = grad_inputs[0]\n\n\n crf = RoICropFunction()\n grid_yx = torch.stack([grid_clone.data[:,:,:,1], grid_clone.data[:,:,:,0]], 3).contiguous().cuda()\n out_stn = crf.forward(input_p, grid_yx)\n grad_inputs = crf.backward(grad_outputs_clone.data)\n grad_input_stn = grad_inputs[0]\n pdb.set_trace()\n\n delta = (grad_input_off.data - grad_input_stn).sum()\n"
},
{
"path": "lib/pycocotools/UPSTREAM_REV",
"content": "https://github.com/pdollar/coco/commit/3ac47c77ebd5a1ed4254a98b7fbf2ef4765a3574\n"
},
{
"path": "lib/pycocotools/__init__.py",
"content": "__author__ = 'tylin'\n"
},
{
"path": "lib/pycocotools/_mask.c",
"content": "/* Generated by Cython 0.29.12 */\n\n#define PY_SSIZE_T_CLEAN\n#include \"Python.h\"\n#ifndef Py_PYTHON_H\n #error Python headers needed to compile C extensions, please install development version of Python.\n#elif PY_VERSION_HEX < 0x02060000 || (0x03000000 <= PY_VERSION_HEX && PY_VERSION_HEX < 0x03030000)\n #error Cython requires Python 2.6+ or Python 3.3+.\n#else\n#define CYTHON_ABI \"0_29_12\"\n#define CYTHON_HEX_VERSION 0x001D0CF0\n#define CYTHON_FUTURE_DIVISION 0\n#include \n#ifndef offsetof\n #define offsetof(type, member) ( (size_t) & ((type*)0) -> member )\n#endif\n#if !defined(WIN32) && !defined(MS_WINDOWS)\n #ifndef __stdcall\n #define __stdcall\n #endif\n #ifndef __cdecl\n #define __cdecl\n #endif\n #ifndef __fastcall\n #define __fastcall\n #endif\n#endif\n#ifndef DL_IMPORT\n #define DL_IMPORT(t) t\n#endif\n#ifndef DL_EXPORT\n #define DL_EXPORT(t) t\n#endif\n#define __PYX_COMMA ,\n#ifndef HAVE_LONG_LONG\n #if PY_VERSION_HEX >= 0x02070000\n #define HAVE_LONG_LONG\n #endif\n#endif\n#ifndef PY_LONG_LONG\n #define PY_LONG_LONG LONG_LONG\n#endif\n#ifndef Py_HUGE_VAL\n #define Py_HUGE_VAL HUGE_VAL\n#endif\n#ifdef PYPY_VERSION\n #define CYTHON_COMPILING_IN_PYPY 1\n #define CYTHON_COMPILING_IN_PYSTON 0\n #define CYTHON_COMPILING_IN_CPYTHON 0\n #undef CYTHON_USE_TYPE_SLOTS\n #define CYTHON_USE_TYPE_SLOTS 0\n #undef CYTHON_USE_PYTYPE_LOOKUP\n #define CYTHON_USE_PYTYPE_LOOKUP 0\n #if PY_VERSION_HEX < 0x03050000\n #undef CYTHON_USE_ASYNC_SLOTS\n #define CYTHON_USE_ASYNC_SLOTS 0\n #elif !defined(CYTHON_USE_ASYNC_SLOTS)\n #define CYTHON_USE_ASYNC_SLOTS 1\n #endif\n #undef CYTHON_USE_PYLIST_INTERNALS\n #define CYTHON_USE_PYLIST_INTERNALS 0\n #undef CYTHON_USE_UNICODE_INTERNALS\n #define CYTHON_USE_UNICODE_INTERNALS 0\n #undef CYTHON_USE_UNICODE_WRITER\n #define CYTHON_USE_UNICODE_WRITER 0\n #undef CYTHON_USE_PYLONG_INTERNALS\n #define CYTHON_USE_PYLONG_INTERNALS 0\n #undef CYTHON_AVOID_BORROWED_REFS\n #define CYTHON_AVOID_BORROWED_REFS 1\n #undef CYTHON_ASSUME_SAFE_MACROS\n #define CYTHON_ASSUME_SAFE_MACROS 0\n #undef CYTHON_UNPACK_METHODS\n #define CYTHON_UNPACK_METHODS 0\n #undef CYTHON_FAST_THREAD_STATE\n #define CYTHON_FAST_THREAD_STATE 0\n #undef CYTHON_FAST_PYCALL\n #define CYTHON_FAST_PYCALL 0\n #undef CYTHON_PEP489_MULTI_PHASE_INIT\n #define CYTHON_PEP489_MULTI_PHASE_INIT 0\n #undef CYTHON_USE_TP_FINALIZE\n #define CYTHON_USE_TP_FINALIZE 0\n #undef CYTHON_USE_DICT_VERSIONS\n #define CYTHON_USE_DICT_VERSIONS 0\n #undef CYTHON_USE_EXC_INFO_STACK\n #define CYTHON_USE_EXC_INFO_STACK 0\n#elif defined(PYSTON_VERSION)\n #define CYTHON_COMPILING_IN_PYPY 0\n #define CYTHON_COMPILING_IN_PYSTON 1\n #define CYTHON_COMPILING_IN_CPYTHON 0\n #ifndef CYTHON_USE_TYPE_SLOTS\n #define CYTHON_USE_TYPE_SLOTS 1\n #endif\n #undef CYTHON_USE_PYTYPE_LOOKUP\n #define CYTHON_USE_PYTYPE_LOOKUP 0\n #undef CYTHON_USE_ASYNC_SLOTS\n #define CYTHON_USE_ASYNC_SLOTS 0\n #undef CYTHON_USE_PYLIST_INTERNALS\n #define CYTHON_USE_PYLIST_INTERNALS 0\n #ifndef CYTHON_USE_UNICODE_INTERNALS\n #define CYTHON_USE_UNICODE_INTERNALS 1\n #endif\n #undef CYTHON_USE_UNICODE_WRITER\n #define CYTHON_USE_UNICODE_WRITER 0\n #undef CYTHON_USE_PYLONG_INTERNALS\n #define CYTHON_USE_PYLONG_INTERNALS 0\n #ifndef CYTHON_AVOID_BORROWED_REFS\n #define CYTHON_AVOID_BORROWED_REFS 0\n #endif\n #ifndef CYTHON_ASSUME_SAFE_MACROS\n #define CYTHON_ASSUME_SAFE_MACROS 1\n #endif\n #ifndef CYTHON_UNPACK_METHODS\n #define CYTHON_UNPACK_METHODS 1\n #endif\n #undef CYTHON_FAST_THREAD_STATE\n #define CYTHON_FAST_THREAD_STATE 0\n #undef CYTHON_FAST_PYCALL\n #define CYTHON_FAST_PYCALL 0\n #undef CYTHON_PEP489_MULTI_PHASE_INIT\n #define CYTHON_PEP489_MULTI_PHASE_INIT 0\n #undef CYTHON_USE_TP_FINALIZE\n #define CYTHON_USE_TP_FINALIZE 0\n #undef CYTHON_USE_DICT_VERSIONS\n #define CYTHON_USE_DICT_VERSIONS 0\n #undef CYTHON_USE_EXC_INFO_STACK\n #define CYTHON_USE_EXC_INFO_STACK 0\n#else\n #define CYTHON_COMPILING_IN_PYPY 0\n #define CYTHON_COMPILING_IN_PYSTON 0\n #define CYTHON_COMPILING_IN_CPYTHON 1\n #ifndef CYTHON_USE_TYPE_SLOTS\n #define CYTHON_USE_TYPE_SLOTS 1\n #endif\n #if PY_VERSION_HEX < 0x02070000\n #undef CYTHON_USE_PYTYPE_LOOKUP\n #define CYTHON_USE_PYTYPE_LOOKUP 0\n #elif !defined(CYTHON_USE_PYTYPE_LOOKUP)\n #define CYTHON_USE_PYTYPE_LOOKUP 1\n #endif\n #if PY_MAJOR_VERSION < 3\n #undef CYTHON_USE_ASYNC_SLOTS\n #define CYTHON_USE_ASYNC_SLOTS 0\n #elif !defined(CYTHON_USE_ASYNC_SLOTS)\n #define CYTHON_USE_ASYNC_SLOTS 1\n #endif\n #if PY_VERSION_HEX < 0x02070000\n #undef CYTHON_USE_PYLONG_INTERNALS\n #define CYTHON_USE_PYLONG_INTERNALS 0\n #elif !defined(CYTHON_USE_PYLONG_INTERNALS)\n #define CYTHON_USE_PYLONG_INTERNALS 1\n #endif\n #ifndef CYTHON_USE_PYLIST_INTERNALS\n #define CYTHON_USE_PYLIST_INTERNALS 1\n #endif\n #ifndef CYTHON_USE_UNICODE_INTERNALS\n #define CYTHON_USE_UNICODE_INTERNALS 1\n #endif\n #if PY_VERSION_HEX < 0x030300F0\n #undef CYTHON_USE_UNICODE_WRITER\n #define CYTHON_USE_UNICODE_WRITER 0\n #elif !defined(CYTHON_USE_UNICODE_WRITER)\n #define CYTHON_USE_UNICODE_WRITER 1\n #endif\n #ifndef CYTHON_AVOID_BORROWED_REFS\n #define CYTHON_AVOID_BORROWED_REFS 0\n #endif\n #ifndef CYTHON_ASSUME_SAFE_MACROS\n #define CYTHON_ASSUME_SAFE_MACROS 1\n #endif\n #ifndef CYTHON_UNPACK_METHODS\n #define CYTHON_UNPACK_METHODS 1\n #endif\n #ifndef CYTHON_FAST_THREAD_STATE\n #define CYTHON_FAST_THREAD_STATE 1\n #endif\n #ifndef CYTHON_FAST_PYCALL\n #define CYTHON_FAST_PYCALL 1\n #endif\n #ifndef CYTHON_PEP489_MULTI_PHASE_INIT\n #define CYTHON_PEP489_MULTI_PHASE_INIT (PY_VERSION_HEX >= 0x03050000)\n #endif\n #ifndef CYTHON_USE_TP_FINALIZE\n #define CYTHON_USE_TP_FINALIZE (PY_VERSION_HEX >= 0x030400a1)\n #endif\n #ifndef CYTHON_USE_DICT_VERSIONS\n #define CYTHON_USE_DICT_VERSIONS (PY_VERSION_HEX >= 0x030600B1)\n #endif\n #ifndef CYTHON_USE_EXC_INFO_STACK\n #define CYTHON_USE_EXC_INFO_STACK (PY_VERSION_HEX >= 0x030700A3)\n #endif\n#endif\n#if !defined(CYTHON_FAST_PYCCALL)\n#define CYTHON_FAST_PYCCALL (CYTHON_FAST_PYCALL && PY_VERSION_HEX >= 0x030600B1)\n#endif\n#if CYTHON_USE_PYLONG_INTERNALS\n #include \"longintrepr.h\"\n #undef SHIFT\n #undef BASE\n #undef MASK\n #ifdef SIZEOF_VOID_P\n enum { __pyx_check_sizeof_voidp = 1 / (int)(SIZEOF_VOID_P == sizeof(void*)) };\n #endif\n#endif\n#ifndef __has_attribute\n #define __has_attribute(x) 0\n#endif\n#ifndef __has_cpp_attribute\n #define __has_cpp_attribute(x) 0\n#endif\n#ifndef CYTHON_RESTRICT\n #if defined(__GNUC__)\n #define CYTHON_RESTRICT __restrict__\n #elif defined(_MSC_VER) && _MSC_VER >= 1400\n #define CYTHON_RESTRICT __restrict\n #elif defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L\n #define CYTHON_RESTRICT restrict\n #else\n #define CYTHON_RESTRICT\n #endif\n#endif\n#ifndef CYTHON_UNUSED\n# if defined(__GNUC__)\n# if !(defined(__cplusplus)) || (__GNUC__ > 3 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 4))\n# define CYTHON_UNUSED __attribute__ ((__unused__))\n# else\n# define CYTHON_UNUSED\n# endif\n# elif defined(__ICC) || (defined(__INTEL_COMPILER) && !defined(_MSC_VER))\n# define CYTHON_UNUSED __attribute__ ((__unused__))\n# else\n# define CYTHON_UNUSED\n# endif\n#endif\n#ifndef CYTHON_MAYBE_UNUSED_VAR\n# if defined(__cplusplus)\n template void CYTHON_MAYBE_UNUSED_VAR( const T& ) { }\n# else\n# define CYTHON_MAYBE_UNUSED_VAR(x) (void)(x)\n# endif\n#endif\n#ifndef CYTHON_NCP_UNUSED\n# if CYTHON_COMPILING_IN_CPYTHON\n# define CYTHON_NCP_UNUSED\n# else\n# define CYTHON_NCP_UNUSED CYTHON_UNUSED\n# endif\n#endif\n#define __Pyx_void_to_None(void_result) ((void)(void_result), Py_INCREF(Py_None), Py_None)\n#ifdef _MSC_VER\n #ifndef _MSC_STDINT_H_\n #if _MSC_VER < 1300\n typedef unsigned char uint8_t;\n typedef unsigned int uint32_t;\n #else\n typedef unsigned __int8 uint8_t;\n typedef unsigned __int32 uint32_t;\n #endif\n #endif\n#else\n #include \n#endif\n#ifndef CYTHON_FALLTHROUGH\n #if defined(__cplusplus) && __cplusplus >= 201103L\n #if __has_cpp_attribute(fallthrough)\n #define CYTHON_FALLTHROUGH [[fallthrough]]\n #elif __has_cpp_attribute(clang::fallthrough)\n #define CYTHON_FALLTHROUGH [[clang::fallthrough]]\n #elif __has_cpp_attribute(gnu::fallthrough)\n #define CYTHON_FALLTHROUGH [[gnu::fallthrough]]\n #endif\n #endif\n #ifndef CYTHON_FALLTHROUGH\n #if __has_attribute(fallthrough)\n #define CYTHON_FALLTHROUGH __attribute__((fallthrough))\n #else\n #define CYTHON_FALLTHROUGH\n #endif\n #endif\n #if defined(__clang__ ) && defined(__apple_build_version__)\n #if __apple_build_version__ < 7000000\n #undef CYTHON_FALLTHROUGH\n #define CYTHON_FALLTHROUGH\n #endif\n #endif\n#endif\n\n#ifndef CYTHON_INLINE\n #if defined(__clang__)\n #define CYTHON_INLINE __inline__ __attribute__ ((__unused__))\n #elif defined(__GNUC__)\n #define CYTHON_INLINE __inline__\n #elif defined(_MSC_VER)\n #define CYTHON_INLINE __inline\n #elif defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L\n #define CYTHON_INLINE inline\n #else\n #define CYTHON_INLINE\n #endif\n#endif\n\n#if CYTHON_COMPILING_IN_PYPY && PY_VERSION_HEX < 0x02070600 && !defined(Py_OptimizeFlag)\n #define Py_OptimizeFlag 0\n#endif\n#define __PYX_BUILD_PY_SSIZE_T \"n\"\n#define CYTHON_FORMAT_SSIZE_T \"z\"\n#if PY_MAJOR_VERSION < 3\n #define __Pyx_BUILTIN_MODULE_NAME \"__builtin__\"\n #define __Pyx_PyCode_New(a, k, l, s, f, code, c, n, v, fv, cell, fn, name, fline, lnos)\\\n PyCode_New(a+k, l, s, f, code, c, n, v, fv, cell, fn, name, fline, lnos)\n #define __Pyx_DefaultClassType PyClass_Type\n#else\n #define __Pyx_BUILTIN_MODULE_NAME \"builtins\"\n#if PY_VERSION_HEX >= 0x030800A4 && PY_VERSION_HEX < 0x030800B2\n #define __Pyx_PyCode_New(a, k, l, s, f, code, c, n, v, fv, cell, fn, name, fline, lnos)\\\n PyCode_New(a, 0, k, l, s, f, code, c, n, v, fv, cell, fn, name, fline, lnos)\n#else\n #define __Pyx_PyCode_New(a, k, l, s, f, code, c, n, v, fv, cell, fn, name, fline, lnos)\\\n PyCode_New(a, k, l, s, f, code, c, n, v, fv, cell, fn, name, fline, lnos)\n#endif\n #define __Pyx_DefaultClassType PyType_Type\n#endif\n#ifndef Py_TPFLAGS_CHECKTYPES\n #define Py_TPFLAGS_CHECKTYPES 0\n#endif\n#ifndef Py_TPFLAGS_HAVE_INDEX\n #define Py_TPFLAGS_HAVE_INDEX 0\n#endif\n#ifndef Py_TPFLAGS_HAVE_NEWBUFFER\n #define Py_TPFLAGS_HAVE_NEWBUFFER 0\n#endif\n#ifndef Py_TPFLAGS_HAVE_FINALIZE\n #define Py_TPFLAGS_HAVE_FINALIZE 0\n#endif\n#ifndef METH_STACKLESS\n #define METH_STACKLESS 0\n#endif\n#if PY_VERSION_HEX <= 0x030700A3 || !defined(METH_FASTCALL)\n #ifndef METH_FASTCALL\n #define METH_FASTCALL 0x80\n #endif\n typedef PyObject *(*__Pyx_PyCFunctionFast) (PyObject *self, PyObject *const *args, Py_ssize_t nargs);\n typedef PyObject *(*__Pyx_PyCFunctionFastWithKeywords) (PyObject *self, PyObject *const *args,\n Py_ssize_t nargs, PyObject *kwnames);\n#else\n #define __Pyx_PyCFunctionFast _PyCFunctionFast\n #define __Pyx_PyCFunctionFastWithKeywords _PyCFunctionFastWithKeywords\n#endif\n#if CYTHON_FAST_PYCCALL\n#define __Pyx_PyFastCFunction_Check(func)\\\n ((PyCFunction_Check(func) && (METH_FASTCALL == (PyCFunction_GET_FLAGS(func) & ~(METH_CLASS | METH_STATIC | METH_COEXIST | METH_KEYWORDS | METH_STACKLESS)))))\n#else\n#define __Pyx_PyFastCFunction_Check(func) 0\n#endif\n#if CYTHON_COMPILING_IN_PYPY && !defined(PyObject_Malloc)\n #define PyObject_Malloc(s) PyMem_Malloc(s)\n #define PyObject_Free(p) PyMem_Free(p)\n #define PyObject_Realloc(p) PyMem_Realloc(p)\n#endif\n#if CYTHON_COMPILING_IN_CPYTHON && PY_VERSION_HEX < 0x030400A1\n #define PyMem_RawMalloc(n) PyMem_Malloc(n)\n #define PyMem_RawRealloc(p, n) PyMem_Realloc(p, n)\n #define PyMem_RawFree(p) PyMem_Free(p)\n#endif\n#if CYTHON_COMPILING_IN_PYSTON\n #define __Pyx_PyCode_HasFreeVars(co) PyCode_HasFreeVars(co)\n #define __Pyx_PyFrame_SetLineNumber(frame, lineno) PyFrame_SetLineNumber(frame, lineno)\n#else\n #define __Pyx_PyCode_HasFreeVars(co) (PyCode_GetNumFree(co) > 0)\n #define __Pyx_PyFrame_SetLineNumber(frame, lineno) (frame)->f_lineno = (lineno)\n#endif\n#if !CYTHON_FAST_THREAD_STATE || PY_VERSION_HEX < 0x02070000\n #define __Pyx_PyThreadState_Current PyThreadState_GET()\n#elif PY_VERSION_HEX >= 0x03060000\n #define __Pyx_PyThreadState_Current _PyThreadState_UncheckedGet()\n#elif PY_VERSION_HEX >= 0x03000000\n #define __Pyx_PyThreadState_Current PyThreadState_GET()\n#else\n #define __Pyx_PyThreadState_Current _PyThreadState_Current\n#endif\n#if PY_VERSION_HEX < 0x030700A2 && !defined(PyThread_tss_create) && !defined(Py_tss_NEEDS_INIT)\n#include \"pythread.h\"\n#define Py_tss_NEEDS_INIT 0\ntypedef int Py_tss_t;\nstatic CYTHON_INLINE int PyThread_tss_create(Py_tss_t *key) {\n *key = PyThread_create_key();\n return 0;\n}\nstatic CYTHON_INLINE Py_tss_t * PyThread_tss_alloc(void) {\n Py_tss_t *key = (Py_tss_t *)PyObject_Malloc(sizeof(Py_tss_t));\n *key = Py_tss_NEEDS_INIT;\n return key;\n}\nstatic CYTHON_INLINE void PyThread_tss_free(Py_tss_t *key) {\n PyObject_Free(key);\n}\nstatic CYTHON_INLINE int PyThread_tss_is_created(Py_tss_t *key) {\n return *key != Py_tss_NEEDS_INIT;\n}\nstatic CYTHON_INLINE void PyThread_tss_delete(Py_tss_t *key) {\n PyThread_delete_key(*key);\n *key = Py_tss_NEEDS_INIT;\n}\nstatic CYTHON_INLINE int PyThread_tss_set(Py_tss_t *key, void *value) {\n return PyThread_set_key_value(*key, value);\n}\nstatic CYTHON_INLINE void * PyThread_tss_get(Py_tss_t *key) {\n return PyThread_get_key_value(*key);\n}\n#endif\n#if CYTHON_COMPILING_IN_CPYTHON || defined(_PyDict_NewPresized)\n#define __Pyx_PyDict_NewPresized(n) ((n <= 8) ? PyDict_New() : _PyDict_NewPresized(n))\n#else\n#define __Pyx_PyDict_NewPresized(n) PyDict_New()\n#endif\n#if PY_MAJOR_VERSION >= 3 || CYTHON_FUTURE_DIVISION\n #define __Pyx_PyNumber_Divide(x,y) PyNumber_TrueDivide(x,y)\n #define __Pyx_PyNumber_InPlaceDivide(x,y) PyNumber_InPlaceTrueDivide(x,y)\n#else\n #define __Pyx_PyNumber_Divide(x,y) PyNumber_Divide(x,y)\n #define __Pyx_PyNumber_InPlaceDivide(x,y) PyNumber_InPlaceDivide(x,y)\n#endif\n#if CYTHON_COMPILING_IN_CPYTHON && PY_VERSION_HEX >= 0x030500A1 && CYTHON_USE_UNICODE_INTERNALS\n#define __Pyx_PyDict_GetItemStr(dict, name) _PyDict_GetItem_KnownHash(dict, name, ((PyASCIIObject *) name)->hash)\n#else\n#define __Pyx_PyDict_GetItemStr(dict, name) PyDict_GetItem(dict, name)\n#endif\n#if PY_VERSION_HEX > 0x03030000 && defined(PyUnicode_KIND)\n #define CYTHON_PEP393_ENABLED 1\n #define __Pyx_PyUnicode_READY(op) (likely(PyUnicode_IS_READY(op)) ?\\\n 0 : _PyUnicode_Ready((PyObject *)(op)))\n #define __Pyx_PyUnicode_GET_LENGTH(u) PyUnicode_GET_LENGTH(u)\n #define __Pyx_PyUnicode_READ_CHAR(u, i) PyUnicode_READ_CHAR(u, i)\n #define __Pyx_PyUnicode_MAX_CHAR_VALUE(u) PyUnicode_MAX_CHAR_VALUE(u)\n #define __Pyx_PyUnicode_KIND(u) PyUnicode_KIND(u)\n #define __Pyx_PyUnicode_DATA(u) PyUnicode_DATA(u)\n #define __Pyx_PyUnicode_READ(k, d, i) PyUnicode_READ(k, d, i)\n #define __Pyx_PyUnicode_WRITE(k, d, i, ch) PyUnicode_WRITE(k, d, i, ch)\n #define __Pyx_PyUnicode_IS_TRUE(u) (0 != (likely(PyUnicode_IS_READY(u)) ? PyUnicode_GET_LENGTH(u) : PyUnicode_GET_SIZE(u)))\n#else\n #define CYTHON_PEP393_ENABLED 0\n #define PyUnicode_1BYTE_KIND 1\n #define PyUnicode_2BYTE_KIND 2\n #define PyUnicode_4BYTE_KIND 4\n #define __Pyx_PyUnicode_READY(op) (0)\n #define __Pyx_PyUnicode_GET_LENGTH(u) PyUnicode_GET_SIZE(u)\n #define __Pyx_PyUnicode_READ_CHAR(u, i) ((Py_UCS4)(PyUnicode_AS_UNICODE(u)[i]))\n #define __Pyx_PyUnicode_MAX_CHAR_VALUE(u) ((sizeof(Py_UNICODE) == 2) ? 65535 : 1114111)\n #define __Pyx_PyUnicode_KIND(u) (sizeof(Py_UNICODE))\n #define __Pyx_PyUnicode_DATA(u) ((void*)PyUnicode_AS_UNICODE(u))\n #define __Pyx_PyUnicode_READ(k, d, i) ((void)(k), (Py_UCS4)(((Py_UNICODE*)d)[i]))\n #define __Pyx_PyUnicode_WRITE(k, d, i, ch) (((void)(k)), ((Py_UNICODE*)d)[i] = ch)\n #define __Pyx_PyUnicode_IS_TRUE(u) (0 != PyUnicode_GET_SIZE(u))\n#endif\n#if CYTHON_COMPILING_IN_PYPY\n #define __Pyx_PyUnicode_Concat(a, b) PyNumber_Add(a, b)\n #define __Pyx_PyUnicode_ConcatSafe(a, b) PyNumber_Add(a, b)\n#else\n #define __Pyx_PyUnicode_Concat(a, b) PyUnicode_Concat(a, b)\n #define __Pyx_PyUnicode_ConcatSafe(a, b) ((unlikely((a) == Py_None) || unlikely((b) == Py_None)) ?\\\n PyNumber_Add(a, b) : __Pyx_PyUnicode_Concat(a, b))\n#endif\n#if CYTHON_COMPILING_IN_PYPY && !defined(PyUnicode_Contains)\n #define PyUnicode_Contains(u, s) PySequence_Contains(u, s)\n#endif\n#if CYTHON_COMPILING_IN_PYPY && !defined(PyByteArray_Check)\n #define PyByteArray_Check(obj) PyObject_TypeCheck(obj, &PyByteArray_Type)\n#endif\n#if CYTHON_COMPILING_IN_PYPY && !defined(PyObject_Format)\n #define PyObject_Format(obj, fmt) PyObject_CallMethod(obj, \"__format__\", \"O\", fmt)\n#endif\n#define __Pyx_PyString_FormatSafe(a, b) ((unlikely((a) == Py_None || (PyString_Check(b) && !PyString_CheckExact(b)))) ? PyNumber_Remainder(a, b) : __Pyx_PyString_Format(a, b))\n#define __Pyx_PyUnicode_FormatSafe(a, b) ((unlikely((a) == Py_None || (PyUnicode_Check(b) && !PyUnicode_CheckExact(b)))) ? PyNumber_Remainder(a, b) : PyUnicode_Format(a, b))\n#if PY_MAJOR_VERSION >= 3\n #define __Pyx_PyString_Format(a, b) PyUnicode_Format(a, b)\n#else\n #define __Pyx_PyString_Format(a, b) PyString_Format(a, b)\n#endif\n#if PY_MAJOR_VERSION < 3 && !defined(PyObject_ASCII)\n #define PyObject_ASCII(o) PyObject_Repr(o)\n#endif\n#if PY_MAJOR_VERSION >= 3\n #define PyBaseString_Type PyUnicode_Type\n #define PyStringObject PyUnicodeObject\n #define PyString_Type PyUnicode_Type\n #define PyString_Check PyUnicode_Check\n #define PyString_CheckExact PyUnicode_CheckExact\n #define PyObject_Unicode PyObject_Str\n#endif\n#if PY_MAJOR_VERSION >= 3\n #define __Pyx_PyBaseString_Check(obj) PyUnicode_Check(obj)\n #define __Pyx_PyBaseString_CheckExact(obj) PyUnicode_CheckExact(obj)\n#else\n #define __Pyx_PyBaseString_Check(obj) (PyString_Check(obj) || PyUnicode_Check(obj))\n #define __Pyx_PyBaseString_CheckExact(obj) (PyString_CheckExact(obj) || PyUnicode_CheckExact(obj))\n#endif\n#ifndef PySet_CheckExact\n #define PySet_CheckExact(obj) (Py_TYPE(obj) == &PySet_Type)\n#endif\n#if CYTHON_ASSUME_SAFE_MACROS\n #define __Pyx_PySequence_SIZE(seq) Py_SIZE(seq)\n#else\n #define __Pyx_PySequence_SIZE(seq) PySequence_Size(seq)\n#endif\n#if PY_MAJOR_VERSION >= 3\n #define PyIntObject PyLongObject\n #define PyInt_Type PyLong_Type\n #define PyInt_Check(op) PyLong_Check(op)\n #define PyInt_CheckExact(op) PyLong_CheckExact(op)\n #define PyInt_FromString PyLong_FromString\n #define PyInt_FromUnicode PyLong_FromUnicode\n #define PyInt_FromLong PyLong_FromLong\n #define PyInt_FromSize_t PyLong_FromSize_t\n #define PyInt_FromSsize_t PyLong_FromSsize_t\n #define PyInt_AsLong PyLong_AsLong\n #define PyInt_AS_LONG PyLong_AS_LONG\n #define PyInt_AsSsize_t PyLong_AsSsize_t\n #define PyInt_AsUnsignedLongMask PyLong_AsUnsignedLongMask\n #define PyInt_AsUnsignedLongLongMask PyLong_AsUnsignedLongLongMask\n #define PyNumber_Int PyNumber_Long\n#endif\n#if PY_MAJOR_VERSION >= 3\n #define PyBoolObject PyLongObject\n#endif\n#if PY_MAJOR_VERSION >= 3 && CYTHON_COMPILING_IN_PYPY\n #ifndef PyUnicode_InternFromString\n #define PyUnicode_InternFromString(s) PyUnicode_FromString(s)\n #endif\n#endif\n#if PY_VERSION_HEX < 0x030200A4\n typedef long Py_hash_t;\n #define __Pyx_PyInt_FromHash_t PyInt_FromLong\n #define __Pyx_PyInt_AsHash_t PyInt_AsLong\n#else\n #define __Pyx_PyInt_FromHash_t PyInt_FromSsize_t\n #define __Pyx_PyInt_AsHash_t PyInt_AsSsize_t\n#endif\n#if PY_MAJOR_VERSION >= 3\n #define __Pyx_PyMethod_New(func, self, klass) ((self) ? PyMethod_New(func, self) : (Py_INCREF(func), func))\n#else\n #define __Pyx_PyMethod_New(func, self, klass) PyMethod_New(func, self, klass)\n#endif\n#if CYTHON_USE_ASYNC_SLOTS\n #if PY_VERSION_HEX >= 0x030500B1\n #define __Pyx_PyAsyncMethodsStruct PyAsyncMethods\n #define __Pyx_PyType_AsAsync(obj) (Py_TYPE(obj)->tp_as_async)\n #else\n #define __Pyx_PyType_AsAsync(obj) ((__Pyx_PyAsyncMethodsStruct*) (Py_TYPE(obj)->tp_reserved))\n #endif\n#else\n #define __Pyx_PyType_AsAsync(obj) NULL\n#endif\n#ifndef __Pyx_PyAsyncMethodsStruct\n typedef struct {\n unaryfunc am_await;\n unaryfunc am_aiter;\n unaryfunc am_anext;\n } __Pyx_PyAsyncMethodsStruct;\n#endif\n\n#if defined(WIN32) || defined(MS_WINDOWS)\n #define _USE_MATH_DEFINES\n#endif\n#include \n#ifdef NAN\n#define __PYX_NAN() ((float) NAN)\n#else\nstatic CYTHON_INLINE float __PYX_NAN() {\n float value;\n memset(&value, 0xFF, sizeof(value));\n return value;\n}\n#endif\n#if defined(__CYGWIN__) && defined(_LDBL_EQ_DBL)\n#define __Pyx_truncl trunc\n#else\n#define __Pyx_truncl truncl\n#endif\n\n\n#define __PYX_ERR(f_index, lineno, Ln_error) \\\n{ \\\n __pyx_filename = __pyx_f[f_index]; __pyx_lineno = lineno; __pyx_clineno = __LINE__; goto Ln_error; \\\n}\n\n#ifndef __PYX_EXTERN_C\n #ifdef __cplusplus\n #define __PYX_EXTERN_C extern \"C\"\n #else\n #define __PYX_EXTERN_C extern\n #endif\n#endif\n\n#define __PYX_HAVE__pycocotools___mask\n#define __PYX_HAVE_API__pycocotools___mask\n/* Early includes */\n#include \n#include \n#include \"numpy/arrayobject.h\"\n#include \"numpy/ufuncobject.h\"\n#include \n#include \"maskApi.h\"\n#ifdef _OPENMP\n#include \n#endif /* _OPENMP */\n\n#if defined(PYREX_WITHOUT_ASSERTIONS) && !defined(CYTHON_WITHOUT_ASSERTIONS)\n#define CYTHON_WITHOUT_ASSERTIONS\n#endif\n\ntypedef struct {PyObject **p; const char *s; const Py_ssize_t n; const char* encoding;\n const char is_unicode; const char is_str; const char intern; } __Pyx_StringTabEntry;\n\n#define __PYX_DEFAULT_STRING_ENCODING_IS_ASCII 0\n#define __PYX_DEFAULT_STRING_ENCODING_IS_UTF8 0\n#define __PYX_DEFAULT_STRING_ENCODING_IS_DEFAULT (PY_MAJOR_VERSION >= 3 && __PYX_DEFAULT_STRING_ENCODING_IS_UTF8)\n#define __PYX_DEFAULT_STRING_ENCODING \"\"\n#define __Pyx_PyObject_FromString __Pyx_PyBytes_FromString\n#define __Pyx_PyObject_FromStringAndSize __Pyx_PyBytes_FromStringAndSize\n#define __Pyx_uchar_cast(c) ((unsigned char)c)\n#define __Pyx_long_cast(x) ((long)x)\n#define __Pyx_fits_Py_ssize_t(v, type, is_signed) (\\\n (sizeof(type) < sizeof(Py_ssize_t)) ||\\\n (sizeof(type) > sizeof(Py_ssize_t) &&\\\n likely(v < (type)PY_SSIZE_T_MAX ||\\\n v == (type)PY_SSIZE_T_MAX) &&\\\n (!is_signed || likely(v > (type)PY_SSIZE_T_MIN ||\\\n v == (type)PY_SSIZE_T_MIN))) ||\\\n (sizeof(type) == sizeof(Py_ssize_t) &&\\\n (is_signed || likely(v < (type)PY_SSIZE_T_MAX ||\\\n v == (type)PY_SSIZE_T_MAX))) )\nstatic CYTHON_INLINE int __Pyx_is_valid_index(Py_ssize_t i, Py_ssize_t limit) {\n return (size_t) i < (size_t) limit;\n}\n#if defined (__cplusplus) && __cplusplus >= 201103L\n #include \n #define __Pyx_sst_abs(value) std::abs(value)\n#elif SIZEOF_INT >= SIZEOF_SIZE_T\n #define __Pyx_sst_abs(value) abs(value)\n#elif SIZEOF_LONG >= SIZEOF_SIZE_T\n #define __Pyx_sst_abs(value) labs(value)\n#elif defined (_MSC_VER)\n #define __Pyx_sst_abs(value) ((Py_ssize_t)_abs64(value))\n#elif defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L\n #define __Pyx_sst_abs(value) llabs(value)\n#elif defined (__GNUC__)\n #define __Pyx_sst_abs(value) __builtin_llabs(value)\n#else\n #define __Pyx_sst_abs(value) ((value<0) ? -value : value)\n#endif\nstatic CYTHON_INLINE const char* __Pyx_PyObject_AsString(PyObject*);\nstatic CYTHON_INLINE const char* __Pyx_PyObject_AsStringAndSize(PyObject*, Py_ssize_t* length);\n#define __Pyx_PyByteArray_FromString(s) PyByteArray_FromStringAndSize((const char*)s, strlen((const char*)s))\n#define __Pyx_PyByteArray_FromStringAndSize(s, l) PyByteArray_FromStringAndSize((const char*)s, l)\n#define __Pyx_PyBytes_FromString PyBytes_FromString\n#define __Pyx_PyBytes_FromStringAndSize PyBytes_FromStringAndSize\nstatic CYTHON_INLINE PyObject* __Pyx_PyUnicode_FromString(const char*);\n#if PY_MAJOR_VERSION < 3\n #define __Pyx_PyStr_FromString __Pyx_PyBytes_FromString\n #define __Pyx_PyStr_FromStringAndSize __Pyx_PyBytes_FromStringAndSize\n#else\n #define __Pyx_PyStr_FromString __Pyx_PyUnicode_FromString\n #define __Pyx_PyStr_FromStringAndSize __Pyx_PyUnicode_FromStringAndSize\n#endif\n#define __Pyx_PyBytes_AsWritableString(s) ((char*) PyBytes_AS_STRING(s))\n#define __Pyx_PyBytes_AsWritableSString(s) ((signed char*) PyBytes_AS_STRING(s))\n#define __Pyx_PyBytes_AsWritableUString(s) ((unsigned char*) PyBytes_AS_STRING(s))\n#define __Pyx_PyBytes_AsString(s) ((const char*) PyBytes_AS_STRING(s))\n#define __Pyx_PyBytes_AsSString(s) ((const signed char*) PyBytes_AS_STRING(s))\n#define __Pyx_PyBytes_AsUString(s) ((const unsigned char*) PyBytes_AS_STRING(s))\n#define __Pyx_PyObject_AsWritableString(s) ((char*) __Pyx_PyObject_AsString(s))\n#define __Pyx_PyObject_AsWritableSString(s) ((signed char*) __Pyx_PyObject_AsString(s))\n#define __Pyx_PyObject_AsWritableUString(s) ((unsigned char*) __Pyx_PyObject_AsString(s))\n#define __Pyx_PyObject_AsSString(s) ((const signed char*) __Pyx_PyObject_AsString(s))\n#define __Pyx_PyObject_AsUString(s) ((const unsigned char*) __Pyx_PyObject_AsString(s))\n#define __Pyx_PyObject_FromCString(s) __Pyx_PyObject_FromString((const char*)s)\n#define __Pyx_PyBytes_FromCString(s) __Pyx_PyBytes_FromString((const char*)s)\n#define __Pyx_PyByteArray_FromCString(s) __Pyx_PyByteArray_FromString((const char*)s)\n#define __Pyx_PyStr_FromCString(s) __Pyx_PyStr_FromString((const char*)s)\n#define __Pyx_PyUnicode_FromCString(s) __Pyx_PyUnicode_FromString((const char*)s)\nstatic CYTHON_INLINE size_t __Pyx_Py_UNICODE_strlen(const Py_UNICODE *u) {\n const Py_UNICODE *u_end = u;\n while (*u_end++) ;\n return (size_t)(u_end - u - 1);\n}\n#define __Pyx_PyUnicode_FromUnicode(u) PyUnicode_FromUnicode(u, __Pyx_Py_UNICODE_strlen(u))\n#define __Pyx_PyUnicode_FromUnicodeAndLength PyUnicode_FromUnicode\n#define __Pyx_PyUnicode_AsUnicode PyUnicode_AsUnicode\n#define __Pyx_NewRef(obj) (Py_INCREF(obj), obj)\n#define __Pyx_Owned_Py_None(b) __Pyx_NewRef(Py_None)\nstatic CYTHON_INLINE PyObject * __Pyx_PyBool_FromLong(long b);\nstatic CYTHON_INLINE int __Pyx_PyObject_IsTrue(PyObject*);\nstatic CYTHON_INLINE int __Pyx_PyObject_IsTrueAndDecref(PyObject*);\nstatic CYTHON_INLINE PyObject* __Pyx_PyNumber_IntOrLong(PyObject* x);\n#define __Pyx_PySequence_Tuple(obj)\\\n (likely(PyTuple_CheckExact(obj)) ? __Pyx_NewRef(obj) : PySequence_Tuple(obj))\nstatic CYTHON_INLINE Py_ssize_t __Pyx_PyIndex_AsSsize_t(PyObject*);\nstatic CYTHON_INLINE PyObject * __Pyx_PyInt_FromSize_t(size_t);\n#if CYTHON_ASSUME_SAFE_MACROS\n#define __pyx_PyFloat_AsDouble(x) (PyFloat_CheckExact(x) ? PyFloat_AS_DOUBLE(x) : PyFloat_AsDouble(x))\n#else\n#define __pyx_PyFloat_AsDouble(x) PyFloat_AsDouble(x)\n#endif\n#define __pyx_PyFloat_AsFloat(x) ((float) __pyx_PyFloat_AsDouble(x))\n#if PY_MAJOR_VERSION >= 3\n#define __Pyx_PyNumber_Int(x) (PyLong_CheckExact(x) ? __Pyx_NewRef(x) : PyNumber_Long(x))\n#else\n#define __Pyx_PyNumber_Int(x) (PyInt_CheckExact(x) ? __Pyx_NewRef(x) : PyNumber_Int(x))\n#endif\n#define __Pyx_PyNumber_Float(x) (PyFloat_CheckExact(x) ? __Pyx_NewRef(x) : PyNumber_Float(x))\n#if PY_MAJOR_VERSION < 3 && __PYX_DEFAULT_STRING_ENCODING_IS_ASCII\nstatic int __Pyx_sys_getdefaultencoding_not_ascii;\nstatic int __Pyx_init_sys_getdefaultencoding_params(void) {\n PyObject* sys;\n PyObject* default_encoding = NULL;\n PyObject* ascii_chars_u = NULL;\n PyObject* ascii_chars_b = NULL;\n const char* default_encoding_c;\n sys = PyImport_ImportModule(\"sys\");\n if (!sys) goto bad;\n default_encoding = PyObject_CallMethod(sys, (char*) \"getdefaultencoding\", NULL);\n Py_DECREF(sys);\n if (!default_encoding) goto bad;\n default_encoding_c = PyBytes_AsString(default_encoding);\n if (!default_encoding_c) goto bad;\n if (strcmp(default_encoding_c, \"ascii\") == 0) {\n __Pyx_sys_getdefaultencoding_not_ascii = 0;\n } else {\n char ascii_chars[128];\n int c;\n for (c = 0; c < 128; c++) {\n ascii_chars[c] = c;\n }\n __Pyx_sys_getdefaultencoding_not_ascii = 1;\n ascii_chars_u = PyUnicode_DecodeASCII(ascii_chars, 128, NULL);\n if (!ascii_chars_u) goto bad;\n ascii_chars_b = PyUnicode_AsEncodedString(ascii_chars_u, default_encoding_c, NULL);\n if (!ascii_chars_b || !PyBytes_Check(ascii_chars_b) || memcmp(ascii_chars, PyBytes_AS_STRING(ascii_chars_b), 128) != 0) {\n PyErr_Format(\n PyExc_ValueError,\n \"This module compiled with c_string_encoding=ascii, but default encoding '%.200s' is not a superset of ascii.\",\n default_encoding_c);\n goto bad;\n }\n Py_DECREF(ascii_chars_u);\n Py_DECREF(ascii_chars_b);\n }\n Py_DECREF(default_encoding);\n return 0;\nbad:\n Py_XDECREF(default_encoding);\n Py_XDECREF(ascii_chars_u);\n Py_XDECREF(ascii_chars_b);\n return -1;\n}\n#endif\n#if __PYX_DEFAULT_STRING_ENCODING_IS_DEFAULT && PY_MAJOR_VERSION >= 3\n#define __Pyx_PyUnicode_FromStringAndSize(c_str, size) PyUnicode_DecodeUTF8(c_str, size, NULL)\n#else\n#define __Pyx_PyUnicode_FromStringAndSize(c_str, size) PyUnicode_Decode(c_str, size, __PYX_DEFAULT_STRING_ENCODING, NULL)\n#if __PYX_DEFAULT_STRING_ENCODING_IS_DEFAULT\nstatic char* __PYX_DEFAULT_STRING_ENCODING;\nstatic int __Pyx_init_sys_getdefaultencoding_params(void) {\n PyObject* sys;\n PyObject* default_encoding = NULL;\n char* default_encoding_c;\n sys = PyImport_ImportModule(\"sys\");\n if (!sys) goto bad;\n default_encoding = PyObject_CallMethod(sys, (char*) (const char*) \"getdefaultencoding\", NULL);\n Py_DECREF(sys);\n if (!default_encoding) goto bad;\n default_encoding_c = PyBytes_AsString(default_encoding);\n if (!default_encoding_c) goto bad;\n __PYX_DEFAULT_STRING_ENCODING = (char*) malloc(strlen(default_encoding_c) + 1);\n if (!__PYX_DEFAULT_STRING_ENCODING) goto bad;\n strcpy(__PYX_DEFAULT_STRING_ENCODING, default_encoding_c);\n Py_DECREF(default_encoding);\n return 0;\nbad:\n Py_XDECREF(default_encoding);\n return -1;\n}\n#endif\n#endif\n\n\n/* Test for GCC > 2.95 */\n#if defined(__GNUC__) && (__GNUC__ > 2 || (__GNUC__ == 2 && (__GNUC_MINOR__ > 95)))\n #define likely(x) __builtin_expect(!!(x), 1)\n #define unlikely(x) __builtin_expect(!!(x), 0)\n#else /* !__GNUC__ or GCC < 2.95 */\n #define likely(x) (x)\n #define unlikely(x) (x)\n#endif /* __GNUC__ */\nstatic CYTHON_INLINE void __Pyx_pretend_to_initialize(void* ptr) { (void)ptr; }\n\nstatic PyObject *__pyx_m = NULL;\nstatic PyObject *__pyx_d;\nstatic PyObject *__pyx_b;\nstatic PyObject *__pyx_cython_runtime = NULL;\nstatic PyObject *__pyx_empty_tuple;\nstatic PyObject *__pyx_empty_bytes;\nstatic PyObject *__pyx_empty_unicode;\nstatic int __pyx_lineno;\nstatic int __pyx_clineno = 0;\nstatic const char * __pyx_cfilenm= __FILE__;\nstatic const char *__pyx_filename;\n\n/* Header.proto */\n#if !defined(CYTHON_CCOMPLEX)\n #if defined(__cplusplus)\n #define CYTHON_CCOMPLEX 1\n #elif defined(_Complex_I)\n #define CYTHON_CCOMPLEX 1\n #else\n #define CYTHON_CCOMPLEX 0\n #endif\n#endif\n#if CYTHON_CCOMPLEX\n #ifdef __cplusplus\n #include \n #else\n #include \n #endif\n#endif\n#if CYTHON_CCOMPLEX && !defined(__cplusplus) && defined(__sun__) && defined(__GNUC__)\n #undef _Complex_I\n #define _Complex_I 1.0fj\n#endif\n\n\nstatic const char *__pyx_f[] = {\n \"pycocotools/_mask.pyx\",\n \"stringsource\",\n \"__init__.pxd\",\n \"type.pxd\",\n};\n/* BufferFormatStructs.proto */\n#define IS_UNSIGNED(type) (((type) -1) > 0)\nstruct __Pyx_StructField_;\n#define __PYX_BUF_FLAGS_PACKED_STRUCT (1 << 0)\ntypedef struct {\n const char* name;\n struct __Pyx_StructField_* fields;\n size_t size;\n size_t arraysize[8];\n int ndim;\n char typegroup;\n char is_unsigned;\n int flags;\n} __Pyx_TypeInfo;\ntypedef struct __Pyx_StructField_ {\n __Pyx_TypeInfo* type;\n const char* name;\n size_t offset;\n} __Pyx_StructField;\ntypedef struct {\n __Pyx_StructField* field;\n size_t parent_offset;\n} __Pyx_BufFmt_StackElem;\ntypedef struct {\n __Pyx_StructField root;\n __Pyx_BufFmt_StackElem* head;\n size_t fmt_offset;\n size_t new_count, enc_count;\n size_t struct_alignment;\n int is_complex;\n char enc_type;\n char new_packmode;\n char enc_packmode;\n char is_valid_array;\n} __Pyx_BufFmt_Context;\n\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":776\n * # in Cython to enable them only on the right systems.\n * \n * ctypedef npy_int8 int8_t # <<<<<<<<<<<<<<\n * ctypedef npy_int16 int16_t\n * ctypedef npy_int32 int32_t\n */\ntypedef npy_int8 __pyx_t_5numpy_int8_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":777\n * \n * ctypedef npy_int8 int8_t\n * ctypedef npy_int16 int16_t # <<<<<<<<<<<<<<\n * ctypedef npy_int32 int32_t\n * ctypedef npy_int64 int64_t\n */\ntypedef npy_int16 __pyx_t_5numpy_int16_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":778\n * ctypedef npy_int8 int8_t\n * ctypedef npy_int16 int16_t\n * ctypedef npy_int32 int32_t # <<<<<<<<<<<<<<\n * ctypedef npy_int64 int64_t\n * #ctypedef npy_int96 int96_t\n */\ntypedef npy_int32 __pyx_t_5numpy_int32_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":779\n * ctypedef npy_int16 int16_t\n * ctypedef npy_int32 int32_t\n * ctypedef npy_int64 int64_t # <<<<<<<<<<<<<<\n * #ctypedef npy_int96 int96_t\n * #ctypedef npy_int128 int128_t\n */\ntypedef npy_int64 __pyx_t_5numpy_int64_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":783\n * #ctypedef npy_int128 int128_t\n * \n * ctypedef npy_uint8 uint8_t # <<<<<<<<<<<<<<\n * ctypedef npy_uint16 uint16_t\n * ctypedef npy_uint32 uint32_t\n */\ntypedef npy_uint8 __pyx_t_5numpy_uint8_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":784\n * \n * ctypedef npy_uint8 uint8_t\n * ctypedef npy_uint16 uint16_t # <<<<<<<<<<<<<<\n * ctypedef npy_uint32 uint32_t\n * ctypedef npy_uint64 uint64_t\n */\ntypedef npy_uint16 __pyx_t_5numpy_uint16_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":785\n * ctypedef npy_uint8 uint8_t\n * ctypedef npy_uint16 uint16_t\n * ctypedef npy_uint32 uint32_t # <<<<<<<<<<<<<<\n * ctypedef npy_uint64 uint64_t\n * #ctypedef npy_uint96 uint96_t\n */\ntypedef npy_uint32 __pyx_t_5numpy_uint32_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":786\n * ctypedef npy_uint16 uint16_t\n * ctypedef npy_uint32 uint32_t\n * ctypedef npy_uint64 uint64_t # <<<<<<<<<<<<<<\n * #ctypedef npy_uint96 uint96_t\n * #ctypedef npy_uint128 uint128_t\n */\ntypedef npy_uint64 __pyx_t_5numpy_uint64_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":790\n * #ctypedef npy_uint128 uint128_t\n * \n * ctypedef npy_float32 float32_t # <<<<<<<<<<<<<<\n * ctypedef npy_float64 float64_t\n * #ctypedef npy_float80 float80_t\n */\ntypedef npy_float32 __pyx_t_5numpy_float32_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":791\n * \n * ctypedef npy_float32 float32_t\n * ctypedef npy_float64 float64_t # <<<<<<<<<<<<<<\n * #ctypedef npy_float80 float80_t\n * #ctypedef npy_float128 float128_t\n */\ntypedef npy_float64 __pyx_t_5numpy_float64_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":800\n * # The int types are mapped a bit surprising --\n * # numpy.int corresponds to 'l' and numpy.long to 'q'\n * ctypedef npy_long int_t # <<<<<<<<<<<<<<\n * ctypedef npy_longlong long_t\n * ctypedef npy_longlong longlong_t\n */\ntypedef npy_long __pyx_t_5numpy_int_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":801\n * # numpy.int corresponds to 'l' and numpy.long to 'q'\n * ctypedef npy_long int_t\n * ctypedef npy_longlong long_t # <<<<<<<<<<<<<<\n * ctypedef npy_longlong longlong_t\n * \n */\ntypedef npy_longlong __pyx_t_5numpy_long_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":802\n * ctypedef npy_long int_t\n * ctypedef npy_longlong long_t\n * ctypedef npy_longlong longlong_t # <<<<<<<<<<<<<<\n * \n * ctypedef npy_ulong uint_t\n */\ntypedef npy_longlong __pyx_t_5numpy_longlong_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":804\n * ctypedef npy_longlong longlong_t\n * \n * ctypedef npy_ulong uint_t # <<<<<<<<<<<<<<\n * ctypedef npy_ulonglong ulong_t\n * ctypedef npy_ulonglong ulonglong_t\n */\ntypedef npy_ulong __pyx_t_5numpy_uint_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":805\n * \n * ctypedef npy_ulong uint_t\n * ctypedef npy_ulonglong ulong_t # <<<<<<<<<<<<<<\n * ctypedef npy_ulonglong ulonglong_t\n * \n */\ntypedef npy_ulonglong __pyx_t_5numpy_ulong_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":806\n * ctypedef npy_ulong uint_t\n * ctypedef npy_ulonglong ulong_t\n * ctypedef npy_ulonglong ulonglong_t # <<<<<<<<<<<<<<\n * \n * ctypedef npy_intp intp_t\n */\ntypedef npy_ulonglong __pyx_t_5numpy_ulonglong_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":808\n * ctypedef npy_ulonglong ulonglong_t\n * \n * ctypedef npy_intp intp_t # <<<<<<<<<<<<<<\n * ctypedef npy_uintp uintp_t\n * \n */\ntypedef npy_intp __pyx_t_5numpy_intp_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":809\n * \n * ctypedef npy_intp intp_t\n * ctypedef npy_uintp uintp_t # <<<<<<<<<<<<<<\n * \n * ctypedef npy_double float_t\n */\ntypedef npy_uintp __pyx_t_5numpy_uintp_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":811\n * ctypedef npy_uintp uintp_t\n * \n * ctypedef npy_double float_t # <<<<<<<<<<<<<<\n * ctypedef npy_double double_t\n * ctypedef npy_longdouble longdouble_t\n */\ntypedef npy_double __pyx_t_5numpy_float_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":812\n * \n * ctypedef npy_double float_t\n * ctypedef npy_double double_t # <<<<<<<<<<<<<<\n * ctypedef npy_longdouble longdouble_t\n * \n */\ntypedef npy_double __pyx_t_5numpy_double_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":813\n * ctypedef npy_double float_t\n * ctypedef npy_double double_t\n * ctypedef npy_longdouble longdouble_t # <<<<<<<<<<<<<<\n * \n * ctypedef npy_cfloat cfloat_t\n */\ntypedef npy_longdouble __pyx_t_5numpy_longdouble_t;\n/* Declarations.proto */\n#if CYTHON_CCOMPLEX\n #ifdef __cplusplus\n typedef ::std::complex< float > __pyx_t_float_complex;\n #else\n typedef float _Complex __pyx_t_float_complex;\n #endif\n#else\n typedef struct { float real, imag; } __pyx_t_float_complex;\n#endif\nstatic CYTHON_INLINE __pyx_t_float_complex __pyx_t_float_complex_from_parts(float, float);\n\n/* Declarations.proto */\n#if CYTHON_CCOMPLEX\n #ifdef __cplusplus\n typedef ::std::complex< double > __pyx_t_double_complex;\n #else\n typedef double _Complex __pyx_t_double_complex;\n #endif\n#else\n typedef struct { double real, imag; } __pyx_t_double_complex;\n#endif\nstatic CYTHON_INLINE __pyx_t_double_complex __pyx_t_double_complex_from_parts(double, double);\n\n\n/*--- Type declarations ---*/\nstruct __pyx_obj_11pycocotools_5_mask_RLEs;\nstruct __pyx_obj_11pycocotools_5_mask_Masks;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":815\n * ctypedef npy_longdouble longdouble_t\n * \n * ctypedef npy_cfloat cfloat_t # <<<<<<<<<<<<<<\n * ctypedef npy_cdouble cdouble_t\n * ctypedef npy_clongdouble clongdouble_t\n */\ntypedef npy_cfloat __pyx_t_5numpy_cfloat_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":816\n * \n * ctypedef npy_cfloat cfloat_t\n * ctypedef npy_cdouble cdouble_t # <<<<<<<<<<<<<<\n * ctypedef npy_clongdouble clongdouble_t\n * \n */\ntypedef npy_cdouble __pyx_t_5numpy_cdouble_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":817\n * ctypedef npy_cfloat cfloat_t\n * ctypedef npy_cdouble cdouble_t\n * ctypedef npy_clongdouble clongdouble_t # <<<<<<<<<<<<<<\n * \n * ctypedef npy_cdouble complex_t\n */\ntypedef npy_clongdouble __pyx_t_5numpy_clongdouble_t;\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":819\n * ctypedef npy_clongdouble clongdouble_t\n * \n * ctypedef npy_cdouble complex_t # <<<<<<<<<<<<<<\n * \n * cdef inline object PyArray_MultiIterNew1(a):\n */\ntypedef npy_cdouble __pyx_t_5numpy_complex_t;\n\n/* \"pycocotools/_mask.pyx\":53\n * # python class to wrap RLE array in C\n * # the class handles the memory allocation and deallocation\n * cdef class RLEs: # <<<<<<<<<<<<<<\n * cdef RLE *_R\n * cdef siz _n\n */\nstruct __pyx_obj_11pycocotools_5_mask_RLEs {\n PyObject_HEAD\n RLE *_R;\n siz _n;\n};\n\n\n/* \"pycocotools/_mask.pyx\":74\n * # python class to wrap Mask array in C\n * # the class handles the memory allocation and deallocation\n * cdef class Masks: # <<<<<<<<<<<<<<\n * cdef byte *_mask\n * cdef siz _h\n */\nstruct __pyx_obj_11pycocotools_5_mask_Masks {\n PyObject_HEAD\n byte *_mask;\n siz _h;\n siz _w;\n siz _n;\n};\n\n\n/* --- Runtime support code (head) --- */\n/* Refnanny.proto */\n#ifndef CYTHON_REFNANNY\n #define CYTHON_REFNANNY 0\n#endif\n#if CYTHON_REFNANNY\n typedef struct {\n void (*INCREF)(void*, PyObject*, int);\n void (*DECREF)(void*, PyObject*, int);\n void (*GOTREF)(void*, PyObject*, int);\n void (*GIVEREF)(void*, PyObject*, int);\n void* (*SetupContext)(const char*, int, const char*);\n void (*FinishContext)(void**);\n } __Pyx_RefNannyAPIStruct;\n static __Pyx_RefNannyAPIStruct *__Pyx_RefNanny = NULL;\n static __Pyx_RefNannyAPIStruct *__Pyx_RefNannyImportAPI(const char *modname);\n #define __Pyx_RefNannyDeclarations void *__pyx_refnanny = NULL;\n#ifdef WITH_THREAD\n #define __Pyx_RefNannySetupContext(name, acquire_gil)\\\n if (acquire_gil) {\\\n PyGILState_STATE __pyx_gilstate_save = PyGILState_Ensure();\\\n __pyx_refnanny = __Pyx_RefNanny->SetupContext((name), __LINE__, __FILE__);\\\n PyGILState_Release(__pyx_gilstate_save);\\\n } else {\\\n __pyx_refnanny = __Pyx_RefNanny->SetupContext((name), __LINE__, __FILE__);\\\n }\n#else\n #define __Pyx_RefNannySetupContext(name, acquire_gil)\\\n __pyx_refnanny = __Pyx_RefNanny->SetupContext((name), __LINE__, __FILE__)\n#endif\n #define __Pyx_RefNannyFinishContext()\\\n __Pyx_RefNanny->FinishContext(&__pyx_refnanny)\n #define __Pyx_INCREF(r) __Pyx_RefNanny->INCREF(__pyx_refnanny, (PyObject *)(r), __LINE__)\n #define __Pyx_DECREF(r) __Pyx_RefNanny->DECREF(__pyx_refnanny, (PyObject *)(r), __LINE__)\n #define __Pyx_GOTREF(r) __Pyx_RefNanny->GOTREF(__pyx_refnanny, (PyObject *)(r), __LINE__)\n #define __Pyx_GIVEREF(r) __Pyx_RefNanny->GIVEREF(__pyx_refnanny, (PyObject *)(r), __LINE__)\n #define __Pyx_XINCREF(r) do { if((r) != NULL) {__Pyx_INCREF(r); }} while(0)\n #define __Pyx_XDECREF(r) do { if((r) != NULL) {__Pyx_DECREF(r); }} while(0)\n #define __Pyx_XGOTREF(r) do { if((r) != NULL) {__Pyx_GOTREF(r); }} while(0)\n #define __Pyx_XGIVEREF(r) do { if((r) != NULL) {__Pyx_GIVEREF(r);}} while(0)\n#else\n #define __Pyx_RefNannyDeclarations\n #define __Pyx_RefNannySetupContext(name, acquire_gil)\n #define __Pyx_RefNannyFinishContext()\n #define __Pyx_INCREF(r) Py_INCREF(r)\n #define __Pyx_DECREF(r) Py_DECREF(r)\n #define __Pyx_GOTREF(r)\n #define __Pyx_GIVEREF(r)\n #define __Pyx_XINCREF(r) Py_XINCREF(r)\n #define __Pyx_XDECREF(r) Py_XDECREF(r)\n #define __Pyx_XGOTREF(r)\n #define __Pyx_XGIVEREF(r)\n#endif\n#define __Pyx_XDECREF_SET(r, v) do {\\\n PyObject *tmp = (PyObject *) r;\\\n r = v; __Pyx_XDECREF(tmp);\\\n } while (0)\n#define __Pyx_DECREF_SET(r, v) do {\\\n PyObject *tmp = (PyObject *) r;\\\n r = v; __Pyx_DECREF(tmp);\\\n } while (0)\n#define __Pyx_CLEAR(r) do { PyObject* tmp = ((PyObject*)(r)); r = NULL; __Pyx_DECREF(tmp);} while(0)\n#define __Pyx_XCLEAR(r) do { if((r) != NULL) {PyObject* tmp = ((PyObject*)(r)); r = NULL; __Pyx_DECREF(tmp);}} while(0)\n\n/* PyObjectGetAttrStr.proto */\n#if CYTHON_USE_TYPE_SLOTS\nstatic CYTHON_INLINE PyObject* __Pyx_PyObject_GetAttrStr(PyObject* obj, PyObject* attr_name);\n#else\n#define __Pyx_PyObject_GetAttrStr(o,n) PyObject_GetAttr(o,n)\n#endif\n\n/* GetBuiltinName.proto */\nstatic PyObject *__Pyx_GetBuiltinName(PyObject *name);\n\n/* RaiseDoubleKeywords.proto */\nstatic void __Pyx_RaiseDoubleKeywordsError(const char* func_name, PyObject* kw_name);\n\n/* ParseKeywords.proto */\nstatic int __Pyx_ParseOptionalKeywords(PyObject *kwds, PyObject **argnames[],\\\n PyObject *kwds2, PyObject *values[], Py_ssize_t num_pos_args,\\\n const char* function_name);\n\n/* RaiseArgTupleInvalid.proto */\nstatic void __Pyx_RaiseArgtupleInvalid(const char* func_name, int exact,\n Py_ssize_t num_min, Py_ssize_t num_max, Py_ssize_t num_found);\n\n/* IncludeStringH.proto */\n#include \n\n/* BytesEquals.proto */\nstatic CYTHON_INLINE int __Pyx_PyBytes_Equals(PyObject* s1, PyObject* s2, int equals);\n\n/* UnicodeEquals.proto */\nstatic CYTHON_INLINE int __Pyx_PyUnicode_Equals(PyObject* s1, PyObject* s2, int equals);\n\n/* StrEquals.proto */\n#if PY_MAJOR_VERSION >= 3\n#define __Pyx_PyString_Equals __Pyx_PyUnicode_Equals\n#else\n#define __Pyx_PyString_Equals __Pyx_PyBytes_Equals\n#endif\n\n/* PyCFunctionFastCall.proto */\n#if CYTHON_FAST_PYCCALL\nstatic CYTHON_INLINE PyObject *__Pyx_PyCFunction_FastCall(PyObject *func, PyObject **args, Py_ssize_t nargs);\n#else\n#define __Pyx_PyCFunction_FastCall(func, args, nargs) (assert(0), NULL)\n#endif\n\n/* PyFunctionFastCall.proto */\n#if CYTHON_FAST_PYCALL\n#define __Pyx_PyFunction_FastCall(func, args, nargs)\\\n __Pyx_PyFunction_FastCallDict((func), (args), (nargs), NULL)\n#if 1 || PY_VERSION_HEX < 0x030600B1\nstatic PyObject *__Pyx_PyFunction_FastCallDict(PyObject *func, PyObject **args, Py_ssize_t nargs, PyObject *kwargs);\n#else\n#define __Pyx_PyFunction_FastCallDict(func, args, nargs, kwargs) _PyFunction_FastCallDict(func, args, nargs, kwargs)\n#endif\n#define __Pyx_BUILD_ASSERT_EXPR(cond)\\\n (sizeof(char [1 - 2*!(cond)]) - 1)\n#ifndef Py_MEMBER_SIZE\n#define Py_MEMBER_SIZE(type, member) sizeof(((type *)0)->member)\n#endif\n static size_t __pyx_pyframe_localsplus_offset = 0;\n #include \"frameobject.h\"\n #define __Pxy_PyFrame_Initialize_Offsets()\\\n ((void)__Pyx_BUILD_ASSERT_EXPR(sizeof(PyFrameObject) == offsetof(PyFrameObject, f_localsplus) + Py_MEMBER_SIZE(PyFrameObject, f_localsplus)),\\\n (void)(__pyx_pyframe_localsplus_offset = ((size_t)PyFrame_Type.tp_basicsize) - Py_MEMBER_SIZE(PyFrameObject, f_localsplus)))\n #define __Pyx_PyFrame_GetLocalsplus(frame)\\\n (assert(__pyx_pyframe_localsplus_offset), (PyObject **)(((char *)(frame)) + __pyx_pyframe_localsplus_offset))\n#endif\n\n/* PyObjectCall.proto */\n#if CYTHON_COMPILING_IN_CPYTHON\nstatic CYTHON_INLINE PyObject* __Pyx_PyObject_Call(PyObject *func, PyObject *arg, PyObject *kw);\n#else\n#define __Pyx_PyObject_Call(func, arg, kw) PyObject_Call(func, arg, kw)\n#endif\n\n/* PyObjectCallMethO.proto */\n#if CYTHON_COMPILING_IN_CPYTHON\nstatic CYTHON_INLINE PyObject* __Pyx_PyObject_CallMethO(PyObject *func, PyObject *arg);\n#endif\n\n/* PyObjectCallOneArg.proto */\nstatic CYTHON_INLINE PyObject* __Pyx_PyObject_CallOneArg(PyObject *func, PyObject *arg);\n\n/* PyThreadStateGet.proto */\n#if CYTHON_FAST_THREAD_STATE\n#define __Pyx_PyThreadState_declare PyThreadState *__pyx_tstate;\n#define __Pyx_PyThreadState_assign __pyx_tstate = __Pyx_PyThreadState_Current;\n#define __Pyx_PyErr_Occurred() __pyx_tstate->curexc_type\n#else\n#define __Pyx_PyThreadState_declare\n#define __Pyx_PyThreadState_assign\n#define __Pyx_PyErr_Occurred() PyErr_Occurred()\n#endif\n\n/* PyErrFetchRestore.proto */\n#if CYTHON_FAST_THREAD_STATE\n#define __Pyx_PyErr_Clear() __Pyx_ErrRestore(NULL, NULL, NULL)\n#define __Pyx_ErrRestoreWithState(type, value, tb) __Pyx_ErrRestoreInState(PyThreadState_GET(), type, value, tb)\n#define __Pyx_ErrFetchWithState(type, value, tb) __Pyx_ErrFetchInState(PyThreadState_GET(), type, value, tb)\n#define __Pyx_ErrRestore(type, value, tb) __Pyx_ErrRestoreInState(__pyx_tstate, type, value, tb)\n#define __Pyx_ErrFetch(type, value, tb) __Pyx_ErrFetchInState(__pyx_tstate, type, value, tb)\nstatic CYTHON_INLINE void __Pyx_ErrRestoreInState(PyThreadState *tstate, PyObject *type, PyObject *value, PyObject *tb);\nstatic CYTHON_INLINE void __Pyx_ErrFetchInState(PyThreadState *tstate, PyObject **type, PyObject **value, PyObject **tb);\n#if CYTHON_COMPILING_IN_CPYTHON\n#define __Pyx_PyErr_SetNone(exc) (Py_INCREF(exc), __Pyx_ErrRestore((exc), NULL, NULL))\n#else\n#define __Pyx_PyErr_SetNone(exc) PyErr_SetNone(exc)\n#endif\n#else\n#define __Pyx_PyErr_Clear() PyErr_Clear()\n#define __Pyx_PyErr_SetNone(exc) PyErr_SetNone(exc)\n#define __Pyx_ErrRestoreWithState(type, value, tb) PyErr_Restore(type, value, tb)\n#define __Pyx_ErrFetchWithState(type, value, tb) PyErr_Fetch(type, value, tb)\n#define __Pyx_ErrRestoreInState(tstate, type, value, tb) PyErr_Restore(type, value, tb)\n#define __Pyx_ErrFetchInState(tstate, type, value, tb) PyErr_Fetch(type, value, tb)\n#define __Pyx_ErrRestore(type, value, tb) PyErr_Restore(type, value, tb)\n#define __Pyx_ErrFetch(type, value, tb) PyErr_Fetch(type, value, tb)\n#endif\n\n/* RaiseException.proto */\nstatic void __Pyx_Raise(PyObject *type, PyObject *value, PyObject *tb, PyObject *cause);\n\n/* ExtTypeTest.proto */\nstatic CYTHON_INLINE int __Pyx_TypeTest(PyObject *obj, PyTypeObject *type);\n\n/* ArgTypeTest.proto */\n#define __Pyx_ArgTypeTest(obj, type, none_allowed, name, exact)\\\n ((likely((Py_TYPE(obj) == type) | (none_allowed && (obj == Py_None)))) ? 1 :\\\n __Pyx__ArgTypeTest(obj, type, name, exact))\nstatic int __Pyx__ArgTypeTest(PyObject *obj, PyTypeObject *type, const char *name, int exact);\n\n/* ListAppend.proto */\n#if CYTHON_USE_PYLIST_INTERNALS && CYTHON_ASSUME_SAFE_MACROS\nstatic CYTHON_INLINE int __Pyx_PyList_Append(PyObject* list, PyObject* x) {\n PyListObject* L = (PyListObject*) list;\n Py_ssize_t len = Py_SIZE(list);\n if (likely(L->allocated > len) & likely(len > (L->allocated >> 1))) {\n Py_INCREF(x);\n PyList_SET_ITEM(list, len, x);\n Py_SIZE(list) = len+1;\n return 0;\n }\n return PyList_Append(list, x);\n}\n#else\n#define __Pyx_PyList_Append(L,x) PyList_Append(L,x)\n#endif\n\n/* PyIntBinop.proto */\n#if !CYTHON_COMPILING_IN_PYPY\nstatic PyObject* __Pyx_PyInt_AddObjC(PyObject *op1, PyObject *op2, long intval, int inplace, int zerodivision_check);\n#else\n#define __Pyx_PyInt_AddObjC(op1, op2, intval, inplace, zerodivision_check)\\\n (inplace ? PyNumber_InPlaceAdd(op1, op2) : PyNumber_Add(op1, op2))\n#endif\n\n/* DictGetItem.proto */\n#if PY_MAJOR_VERSION >= 3 && !CYTHON_COMPILING_IN_PYPY\nstatic PyObject *__Pyx_PyDict_GetItem(PyObject *d, PyObject* key);\n#define __Pyx_PyObject_Dict_GetItem(obj, name)\\\n (likely(PyDict_CheckExact(obj)) ?\\\n __Pyx_PyDict_GetItem(obj, name) : PyObject_GetItem(obj, name))\n#else\n#define __Pyx_PyDict_GetItem(d, key) PyObject_GetItem(d, key)\n#define __Pyx_PyObject_Dict_GetItem(obj, name) PyObject_GetItem(obj, name)\n#endif\n\n/* GetItemInt.proto */\n#define __Pyx_GetItemInt(o, i, type, is_signed, to_py_func, is_list, wraparound, boundscheck)\\\n (__Pyx_fits_Py_ssize_t(i, type, is_signed) ?\\\n __Pyx_GetItemInt_Fast(o, (Py_ssize_t)i, is_list, wraparound, boundscheck) :\\\n (is_list ? (PyErr_SetString(PyExc_IndexError, \"list index out of range\"), (PyObject*)NULL) :\\\n __Pyx_GetItemInt_Generic(o, to_py_func(i))))\n#define __Pyx_GetItemInt_List(o, i, type, is_signed, to_py_func, is_list, wraparound, boundscheck)\\\n (__Pyx_fits_Py_ssize_t(i, type, is_signed) ?\\\n __Pyx_GetItemInt_List_Fast(o, (Py_ssize_t)i, wraparound, boundscheck) :\\\n (PyErr_SetString(PyExc_IndexError, \"list index out of range\"), (PyObject*)NULL))\nstatic CYTHON_INLINE PyObject *__Pyx_GetItemInt_List_Fast(PyObject *o, Py_ssize_t i,\n int wraparound, int boundscheck);\n#define __Pyx_GetItemInt_Tuple(o, i, type, is_signed, to_py_func, is_list, wraparound, boundscheck)\\\n (__Pyx_fits_Py_ssize_t(i, type, is_signed) ?\\\n __Pyx_GetItemInt_Tuple_Fast(o, (Py_ssize_t)i, wraparound, boundscheck) :\\\n (PyErr_SetString(PyExc_IndexError, \"tuple index out of range\"), (PyObject*)NULL))\nstatic CYTHON_INLINE PyObject *__Pyx_GetItemInt_Tuple_Fast(PyObject *o, Py_ssize_t i,\n int wraparound, int boundscheck);\nstatic PyObject *__Pyx_GetItemInt_Generic(PyObject *o, PyObject* j);\nstatic CYTHON_INLINE PyObject *__Pyx_GetItemInt_Fast(PyObject *o, Py_ssize_t i,\n int is_list, int wraparound, int boundscheck);\n\n/* IsLittleEndian.proto */\nstatic CYTHON_INLINE int __Pyx_Is_Little_Endian(void);\n\n/* BufferFormatCheck.proto */\nstatic const char* __Pyx_BufFmt_CheckString(__Pyx_BufFmt_Context* ctx, const char* ts);\nstatic void __Pyx_BufFmt_Init(__Pyx_BufFmt_Context* ctx,\n __Pyx_BufFmt_StackElem* stack,\n __Pyx_TypeInfo* type);\n\n/* BufferGetAndValidate.proto */\n#define __Pyx_GetBufferAndValidate(buf, obj, dtype, flags, nd, cast, stack)\\\n ((obj == Py_None || obj == NULL) ?\\\n (__Pyx_ZeroBuffer(buf), 0) :\\\n __Pyx__GetBufferAndValidate(buf, obj, dtype, flags, nd, cast, stack))\nstatic int __Pyx__GetBufferAndValidate(Py_buffer* buf, PyObject* obj,\n __Pyx_TypeInfo* dtype, int flags, int nd, int cast, __Pyx_BufFmt_StackElem* stack);\nstatic void __Pyx_ZeroBuffer(Py_buffer* buf);\nstatic CYTHON_INLINE void __Pyx_SafeReleaseBuffer(Py_buffer* info);\nstatic Py_ssize_t __Pyx_minusones[] = { -1, -1, -1, -1, -1, -1, -1, -1 };\nstatic Py_ssize_t __Pyx_zeros[] = { 0, 0, 0, 0, 0, 0, 0, 0 };\n\n/* PyDictVersioning.proto */\n#if CYTHON_USE_DICT_VERSIONS && CYTHON_USE_TYPE_SLOTS\n#define __PYX_DICT_VERSION_INIT ((PY_UINT64_T) -1)\n#define __PYX_GET_DICT_VERSION(dict) (((PyDictObject*)(dict))->ma_version_tag)\n#define __PYX_UPDATE_DICT_CACHE(dict, value, cache_var, version_var)\\\n (version_var) = __PYX_GET_DICT_VERSION(dict);\\\n (cache_var) = (value);\n#define __PYX_PY_DICT_LOOKUP_IF_MODIFIED(VAR, DICT, LOOKUP) {\\\n static PY_UINT64_T __pyx_dict_version = 0;\\\n static PyObject *__pyx_dict_cached_value = NULL;\\\n if (likely(__PYX_GET_DICT_VERSION(DICT) == __pyx_dict_version)) {\\\n (VAR) = __pyx_dict_cached_value;\\\n } else {\\\n (VAR) = __pyx_dict_cached_value = (LOOKUP);\\\n __pyx_dict_version = __PYX_GET_DICT_VERSION(DICT);\\\n }\\\n}\nstatic CYTHON_INLINE PY_UINT64_T __Pyx_get_tp_dict_version(PyObject *obj);\nstatic CYTHON_INLINE PY_UINT64_T __Pyx_get_object_dict_version(PyObject *obj);\nstatic CYTHON_INLINE int __Pyx_object_dict_version_matches(PyObject* obj, PY_UINT64_T tp_dict_version, PY_UINT64_T obj_dict_version);\n#else\n#define __PYX_GET_DICT_VERSION(dict) (0)\n#define __PYX_UPDATE_DICT_CACHE(dict, value, cache_var, version_var)\n#define __PYX_PY_DICT_LOOKUP_IF_MODIFIED(VAR, DICT, LOOKUP) (VAR) = (LOOKUP);\n#endif\n\n/* GetModuleGlobalName.proto */\n#if CYTHON_USE_DICT_VERSIONS\n#define __Pyx_GetModuleGlobalName(var, name) {\\\n static PY_UINT64_T __pyx_dict_version = 0;\\\n static PyObject *__pyx_dict_cached_value = NULL;\\\n (var) = (likely(__pyx_dict_version == __PYX_GET_DICT_VERSION(__pyx_d))) ?\\\n (likely(__pyx_dict_cached_value) ? __Pyx_NewRef(__pyx_dict_cached_value) : __Pyx_GetBuiltinName(name)) :\\\n __Pyx__GetModuleGlobalName(name, &__pyx_dict_version, &__pyx_dict_cached_value);\\\n}\n#define __Pyx_GetModuleGlobalNameUncached(var, name) {\\\n PY_UINT64_T __pyx_dict_version;\\\n PyObject *__pyx_dict_cached_value;\\\n (var) = __Pyx__GetModuleGlobalName(name, &__pyx_dict_version, &__pyx_dict_cached_value);\\\n}\nstatic PyObject *__Pyx__GetModuleGlobalName(PyObject *name, PY_UINT64_T *dict_version, PyObject **dict_cached_value);\n#else\n#define __Pyx_GetModuleGlobalName(var, name) (var) = __Pyx__GetModuleGlobalName(name)\n#define __Pyx_GetModuleGlobalNameUncached(var, name) (var) = __Pyx__GetModuleGlobalName(name)\nstatic CYTHON_INLINE PyObject *__Pyx__GetModuleGlobalName(PyObject *name);\n#endif\n\n/* PyObjectCall2Args.proto */\nstatic CYTHON_UNUSED PyObject* __Pyx_PyObject_Call2Args(PyObject* function, PyObject* arg1, PyObject* arg2);\n\n/* PyIntCompare.proto */\nstatic CYTHON_INLINE PyObject* __Pyx_PyInt_EqObjC(PyObject *op1, PyObject *op2, long intval, long inplace);\n\n/* ListCompAppend.proto */\n#if CYTHON_USE_PYLIST_INTERNALS && CYTHON_ASSUME_SAFE_MACROS\nstatic CYTHON_INLINE int __Pyx_ListComp_Append(PyObject* list, PyObject* x) {\n PyListObject* L = (PyListObject*) list;\n Py_ssize_t len = Py_SIZE(list);\n if (likely(L->allocated > len)) {\n Py_INCREF(x);\n PyList_SET_ITEM(list, len, x);\n Py_SIZE(list) = len+1;\n return 0;\n }\n return PyList_Append(list, x);\n}\n#else\n#define __Pyx_ListComp_Append(L,x) PyList_Append(L,x)\n#endif\n\n/* FetchCommonType.proto */\nstatic PyTypeObject* __Pyx_FetchCommonType(PyTypeObject* type);\n\n/* CythonFunction.proto */\n#define __Pyx_CyFunction_USED 1\n#define __Pyx_CYFUNCTION_STATICMETHOD 0x01\n#define __Pyx_CYFUNCTION_CLASSMETHOD 0x02\n#define __Pyx_CYFUNCTION_CCLASS 0x04\n#define __Pyx_CyFunction_GetClosure(f)\\\n (((__pyx_CyFunctionObject *) (f))->func_closure)\n#define __Pyx_CyFunction_GetClassObj(f)\\\n (((__pyx_CyFunctionObject *) (f))->func_classobj)\n#define __Pyx_CyFunction_Defaults(type, f)\\\n ((type *)(((__pyx_CyFunctionObject *) (f))->defaults))\n#define __Pyx_CyFunction_SetDefaultsGetter(f, g)\\\n ((__pyx_CyFunctionObject *) (f))->defaults_getter = (g)\ntypedef struct {\n PyCFunctionObject func;\n#if PY_VERSION_HEX < 0x030500A0\n PyObject *func_weakreflist;\n#endif\n PyObject *func_dict;\n PyObject *func_name;\n PyObject *func_qualname;\n PyObject *func_doc;\n PyObject *func_globals;\n PyObject *func_code;\n PyObject *func_closure;\n PyObject *func_classobj;\n void *defaults;\n int defaults_pyobjects;\n int flags;\n PyObject *defaults_tuple;\n PyObject *defaults_kwdict;\n PyObject *(*defaults_getter)(PyObject *);\n PyObject *func_annotations;\n} __pyx_CyFunctionObject;\nstatic PyTypeObject *__pyx_CyFunctionType = 0;\n#define __Pyx_CyFunction_Check(obj) (__Pyx_TypeCheck(obj, __pyx_CyFunctionType))\n#define __Pyx_CyFunction_NewEx(ml, flags, qualname, self, module, globals, code)\\\n __Pyx_CyFunction_New(__pyx_CyFunctionType, ml, flags, qualname, self, module, globals, code)\nstatic PyObject *__Pyx_CyFunction_New(PyTypeObject *, PyMethodDef *ml,\n int flags, PyObject* qualname,\n PyObject *self,\n PyObject *module, PyObject *globals,\n PyObject* code);\nstatic CYTHON_INLINE void *__Pyx_CyFunction_InitDefaults(PyObject *m,\n size_t size,\n int pyobjects);\nstatic CYTHON_INLINE void __Pyx_CyFunction_SetDefaultsTuple(PyObject *m,\n PyObject *tuple);\nstatic CYTHON_INLINE void __Pyx_CyFunction_SetDefaultsKwDict(PyObject *m,\n PyObject *dict);\nstatic CYTHON_INLINE void __Pyx_CyFunction_SetAnnotationsDict(PyObject *m,\n PyObject *dict);\nstatic int __pyx_CyFunction_init(void);\n\n/* BufferFallbackError.proto */\nstatic void __Pyx_RaiseBufferFallbackError(void);\n\n/* None.proto */\nstatic CYTHON_INLINE Py_ssize_t __Pyx_div_Py_ssize_t(Py_ssize_t, Py_ssize_t);\n\n/* BufferIndexError.proto */\nstatic void __Pyx_RaiseBufferIndexError(int axis);\n\n#define __Pyx_BufPtrStrided1d(type, buf, i0, s0) (type)((char*)buf + i0 * s0)\n/* RaiseTooManyValuesToUnpack.proto */\nstatic CYTHON_INLINE void __Pyx_RaiseTooManyValuesError(Py_ssize_t expected);\n\n/* RaiseNeedMoreValuesToUnpack.proto */\nstatic CYTHON_INLINE void __Pyx_RaiseNeedMoreValuesError(Py_ssize_t index);\n\n/* RaiseNoneIterError.proto */\nstatic CYTHON_INLINE void __Pyx_RaiseNoneNotIterableError(void);\n\n/* GetTopmostException.proto */\n#if CYTHON_USE_EXC_INFO_STACK\nstatic _PyErr_StackItem * __Pyx_PyErr_GetTopmostException(PyThreadState *tstate);\n#endif\n\n/* SaveResetException.proto */\n#if CYTHON_FAST_THREAD_STATE\n#define __Pyx_ExceptionSave(type, value, tb) __Pyx__ExceptionSave(__pyx_tstate, type, value, tb)\nstatic CYTHON_INLINE void __Pyx__ExceptionSave(PyThreadState *tstate, PyObject **type, PyObject **value, PyObject **tb);\n#define __Pyx_ExceptionReset(type, value, tb) __Pyx__ExceptionReset(__pyx_tstate, type, value, tb)\nstatic CYTHON_INLINE void __Pyx__ExceptionReset(PyThreadState *tstate, PyObject *type, PyObject *value, PyObject *tb);\n#else\n#define __Pyx_ExceptionSave(type, value, tb) PyErr_GetExcInfo(type, value, tb)\n#define __Pyx_ExceptionReset(type, value, tb) PyErr_SetExcInfo(type, value, tb)\n#endif\n\n/* PyErrExceptionMatches.proto */\n#if CYTHON_FAST_THREAD_STATE\n#define __Pyx_PyErr_ExceptionMatches(err) __Pyx_PyErr_ExceptionMatchesInState(__pyx_tstate, err)\nstatic CYTHON_INLINE int __Pyx_PyErr_ExceptionMatchesInState(PyThreadState* tstate, PyObject* err);\n#else\n#define __Pyx_PyErr_ExceptionMatches(err) PyErr_ExceptionMatches(err)\n#endif\n\n/* GetException.proto */\n#if CYTHON_FAST_THREAD_STATE\n#define __Pyx_GetException(type, value, tb) __Pyx__GetException(__pyx_tstate, type, value, tb)\nstatic int __Pyx__GetException(PyThreadState *tstate, PyObject **type, PyObject **value, PyObject **tb);\n#else\nstatic int __Pyx_GetException(PyObject **type, PyObject **value, PyObject **tb);\n#endif\n\n/* PyObject_GenericGetAttrNoDict.proto */\n#if CYTHON_USE_TYPE_SLOTS && CYTHON_USE_PYTYPE_LOOKUP && PY_VERSION_HEX < 0x03070000\nstatic CYTHON_INLINE PyObject* __Pyx_PyObject_GenericGetAttrNoDict(PyObject* obj, PyObject* attr_name);\n#else\n#define __Pyx_PyObject_GenericGetAttrNoDict PyObject_GenericGetAttr\n#endif\n\n/* PyObject_GenericGetAttr.proto */\n#if CYTHON_USE_TYPE_SLOTS && CYTHON_USE_PYTYPE_LOOKUP && PY_VERSION_HEX < 0x03070000\nstatic PyObject* __Pyx_PyObject_GenericGetAttr(PyObject* obj, PyObject* attr_name);\n#else\n#define __Pyx_PyObject_GenericGetAttr PyObject_GenericGetAttr\n#endif\n\n/* SetupReduce.proto */\nstatic int __Pyx_setup_reduce(PyObject* type_obj);\n\n/* TypeImport.proto */\n#ifndef __PYX_HAVE_RT_ImportType_proto\n#define __PYX_HAVE_RT_ImportType_proto\nenum __Pyx_ImportType_CheckSize {\n __Pyx_ImportType_CheckSize_Error = 0,\n __Pyx_ImportType_CheckSize_Warn = 1,\n __Pyx_ImportType_CheckSize_Ignore = 2\n};\nstatic PyTypeObject *__Pyx_ImportType(PyObject* module, const char *module_name, const char *class_name, size_t size, enum __Pyx_ImportType_CheckSize check_size);\n#endif\n\n/* Import.proto */\nstatic PyObject *__Pyx_Import(PyObject *name, PyObject *from_list, int level);\n\n/* CLineInTraceback.proto */\n#ifdef CYTHON_CLINE_IN_TRACEBACK\n#define __Pyx_CLineForTraceback(tstate, c_line) (((CYTHON_CLINE_IN_TRACEBACK)) ? c_line : 0)\n#else\nstatic int __Pyx_CLineForTraceback(PyThreadState *tstate, int c_line);\n#endif\n\n/* CodeObjectCache.proto */\ntypedef struct {\n PyCodeObject* code_object;\n int code_line;\n} __Pyx_CodeObjectCacheEntry;\nstruct __Pyx_CodeObjectCache {\n int count;\n int max_count;\n __Pyx_CodeObjectCacheEntry* entries;\n};\nstatic struct __Pyx_CodeObjectCache __pyx_code_cache = {0,0,NULL};\nstatic int __pyx_bisect_code_objects(__Pyx_CodeObjectCacheEntry* entries, int count, int code_line);\nstatic PyCodeObject *__pyx_find_code_object(int code_line);\nstatic void __pyx_insert_code_object(int code_line, PyCodeObject* code_object);\n\n/* AddTraceback.proto */\nstatic void __Pyx_AddTraceback(const char *funcname, int c_line,\n int py_line, const char *filename);\n\n/* BufferStructDeclare.proto */\ntypedef struct {\n Py_ssize_t shape, strides, suboffsets;\n} __Pyx_Buf_DimInfo;\ntypedef struct {\n size_t refcount;\n Py_buffer pybuffer;\n} __Pyx_Buffer;\ntypedef struct {\n __Pyx_Buffer *rcbuffer;\n char *data;\n __Pyx_Buf_DimInfo diminfo[8];\n} __Pyx_LocalBuf_ND;\n\n#if PY_MAJOR_VERSION < 3\n static int __Pyx_GetBuffer(PyObject *obj, Py_buffer *view, int flags);\n static void __Pyx_ReleaseBuffer(Py_buffer *view);\n#else\n #define __Pyx_GetBuffer PyObject_GetBuffer\n #define __Pyx_ReleaseBuffer PyBuffer_Release\n#endif\n\n\n/* CIntToPy.proto */\nstatic CYTHON_INLINE PyObject* __Pyx_PyInt_From_siz(siz value);\n\n/* CIntToPy.proto */\nstatic CYTHON_INLINE PyObject* __Pyx_PyInt_From_long(long value);\n\n/* CIntToPy.proto */\nstatic CYTHON_INLINE PyObject* __Pyx_PyInt_From_Py_intptr_t(Py_intptr_t value);\n\n/* RealImag.proto */\n#if CYTHON_CCOMPLEX\n #ifdef __cplusplus\n #define __Pyx_CREAL(z) ((z).real())\n #define __Pyx_CIMAG(z) ((z).imag())\n #else\n #define __Pyx_CREAL(z) (__real__(z))\n #define __Pyx_CIMAG(z) (__imag__(z))\n #endif\n#else\n #define __Pyx_CREAL(z) ((z).real)\n #define __Pyx_CIMAG(z) ((z).imag)\n#endif\n#if defined(__cplusplus) && CYTHON_CCOMPLEX\\\n && (defined(_WIN32) || defined(__clang__) || (defined(__GNUC__) && (__GNUC__ >= 5 || __GNUC__ == 4 && __GNUC_MINOR__ >= 4 )) || __cplusplus >= 201103)\n #define __Pyx_SET_CREAL(z,x) ((z).real(x))\n #define __Pyx_SET_CIMAG(z,y) ((z).imag(y))\n#else\n #define __Pyx_SET_CREAL(z,x) __Pyx_CREAL(z) = (x)\n #define __Pyx_SET_CIMAG(z,y) __Pyx_CIMAG(z) = (y)\n#endif\n\n/* Arithmetic.proto */\n#if CYTHON_CCOMPLEX\n #define __Pyx_c_eq_float(a, b) ((a)==(b))\n #define __Pyx_c_sum_float(a, b) ((a)+(b))\n #define __Pyx_c_diff_float(a, b) ((a)-(b))\n #define __Pyx_c_prod_float(a, b) ((a)*(b))\n #define __Pyx_c_quot_float(a, b) ((a)/(b))\n #define __Pyx_c_neg_float(a) (-(a))\n #ifdef __cplusplus\n #define __Pyx_c_is_zero_float(z) ((z)==(float)0)\n #define __Pyx_c_conj_float(z) (::std::conj(z))\n #if 1\n #define __Pyx_c_abs_float(z) (::std::abs(z))\n #define __Pyx_c_pow_float(a, b) (::std::pow(a, b))\n #endif\n #else\n #define __Pyx_c_is_zero_float(z) ((z)==0)\n #define __Pyx_c_conj_float(z) (conjf(z))\n #if 1\n #define __Pyx_c_abs_float(z) (cabsf(z))\n #define __Pyx_c_pow_float(a, b) (cpowf(a, b))\n #endif\n #endif\n#else\n static CYTHON_INLINE int __Pyx_c_eq_float(__pyx_t_float_complex, __pyx_t_float_complex);\n static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_sum_float(__pyx_t_float_complex, __pyx_t_float_complex);\n static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_diff_float(__pyx_t_float_complex, __pyx_t_float_complex);\n static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_prod_float(__pyx_t_float_complex, __pyx_t_float_complex);\n static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_quot_float(__pyx_t_float_complex, __pyx_t_float_complex);\n static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_neg_float(__pyx_t_float_complex);\n static CYTHON_INLINE int __Pyx_c_is_zero_float(__pyx_t_float_complex);\n static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_conj_float(__pyx_t_float_complex);\n #if 1\n static CYTHON_INLINE float __Pyx_c_abs_float(__pyx_t_float_complex);\n static CYTHON_INLINE __pyx_t_float_complex __Pyx_c_pow_float(__pyx_t_float_complex, __pyx_t_float_complex);\n #endif\n#endif\n\n/* Arithmetic.proto */\n#if CYTHON_CCOMPLEX\n #define __Pyx_c_eq_double(a, b) ((a)==(b))\n #define __Pyx_c_sum_double(a, b) ((a)+(b))\n #define __Pyx_c_diff_double(a, b) ((a)-(b))\n #define __Pyx_c_prod_double(a, b) ((a)*(b))\n #define __Pyx_c_quot_double(a, b) ((a)/(b))\n #define __Pyx_c_neg_double(a) (-(a))\n #ifdef __cplusplus\n #define __Pyx_c_is_zero_double(z) ((z)==(double)0)\n #define __Pyx_c_conj_double(z) (::std::conj(z))\n #if 1\n #define __Pyx_c_abs_double(z) (::std::abs(z))\n #define __Pyx_c_pow_double(a, b) (::std::pow(a, b))\n #endif\n #else\n #define __Pyx_c_is_zero_double(z) ((z)==0)\n #define __Pyx_c_conj_double(z) (conj(z))\n #if 1\n #define __Pyx_c_abs_double(z) (cabs(z))\n #define __Pyx_c_pow_double(a, b) (cpow(a, b))\n #endif\n #endif\n#else\n static CYTHON_INLINE int __Pyx_c_eq_double(__pyx_t_double_complex, __pyx_t_double_complex);\n static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_sum_double(__pyx_t_double_complex, __pyx_t_double_complex);\n static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_diff_double(__pyx_t_double_complex, __pyx_t_double_complex);\n static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_prod_double(__pyx_t_double_complex, __pyx_t_double_complex);\n static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_quot_double(__pyx_t_double_complex, __pyx_t_double_complex);\n static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_neg_double(__pyx_t_double_complex);\n static CYTHON_INLINE int __Pyx_c_is_zero_double(__pyx_t_double_complex);\n static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_conj_double(__pyx_t_double_complex);\n #if 1\n static CYTHON_INLINE double __Pyx_c_abs_double(__pyx_t_double_complex);\n static CYTHON_INLINE __pyx_t_double_complex __Pyx_c_pow_double(__pyx_t_double_complex, __pyx_t_double_complex);\n #endif\n#endif\n\n/* CIntToPy.proto */\nstatic CYTHON_INLINE PyObject* __Pyx_PyInt_From_int(int value);\n\n/* CIntToPy.proto */\nstatic CYTHON_INLINE PyObject* __Pyx_PyInt_From_enum__NPY_TYPES(enum NPY_TYPES value);\n\n/* CIntFromPy.proto */\nstatic CYTHON_INLINE siz __Pyx_PyInt_As_siz(PyObject *);\n\n/* CIntFromPy.proto */\nstatic CYTHON_INLINE size_t __Pyx_PyInt_As_size_t(PyObject *);\n\n/* CIntFromPy.proto */\nstatic CYTHON_INLINE int __Pyx_PyInt_As_int(PyObject *);\n\n/* CIntFromPy.proto */\nstatic CYTHON_INLINE long __Pyx_PyInt_As_long(PyObject *);\n\n/* FastTypeChecks.proto */\n#if CYTHON_COMPILING_IN_CPYTHON\n#define __Pyx_TypeCheck(obj, type) __Pyx_IsSubtype(Py_TYPE(obj), (PyTypeObject *)type)\nstatic CYTHON_INLINE int __Pyx_IsSubtype(PyTypeObject *a, PyTypeObject *b);\nstatic CYTHON_INLINE int __Pyx_PyErr_GivenExceptionMatches(PyObject *err, PyObject *type);\nstatic CYTHON_INLINE int __Pyx_PyErr_GivenExceptionMatches2(PyObject *err, PyObject *type1, PyObject *type2);\n#else\n#define __Pyx_TypeCheck(obj, type) PyObject_TypeCheck(obj, (PyTypeObject *)type)\n#define __Pyx_PyErr_GivenExceptionMatches(err, type) PyErr_GivenExceptionMatches(err, type)\n#define __Pyx_PyErr_GivenExceptionMatches2(err, type1, type2) (PyErr_GivenExceptionMatches(err, type1) || PyErr_GivenExceptionMatches(err, type2))\n#endif\n#define __Pyx_PyException_Check(obj) __Pyx_TypeCheck(obj, PyExc_Exception)\n\n/* CheckBinaryVersion.proto */\nstatic int __Pyx_check_binary_version(void);\n\n/* InitStrings.proto */\nstatic int __Pyx_InitStrings(__Pyx_StringTabEntry *t);\n\n\n/* Module declarations from 'cpython.buffer' */\n\n/* Module declarations from 'libc.string' */\n\n/* Module declarations from 'libc.stdio' */\n\n/* Module declarations from '__builtin__' */\n\n/* Module declarations from 'cpython.type' */\nstatic PyTypeObject *__pyx_ptype_7cpython_4type_type = 0;\n\n/* Module declarations from 'cpython' */\n\n/* Module declarations from 'cpython.object' */\n\n/* Module declarations from 'cpython.ref' */\n\n/* Module declarations from 'cpython.mem' */\n\n/* Module declarations from 'numpy' */\n\n/* Module declarations from 'numpy' */\nstatic PyTypeObject *__pyx_ptype_5numpy_dtype = 0;\nstatic PyTypeObject *__pyx_ptype_5numpy_flatiter = 0;\nstatic PyTypeObject *__pyx_ptype_5numpy_broadcast = 0;\nstatic PyTypeObject *__pyx_ptype_5numpy_ndarray = 0;\nstatic PyTypeObject *__pyx_ptype_5numpy_ufunc = 0;\nstatic CYTHON_INLINE char *__pyx_f_5numpy__util_dtypestring(PyArray_Descr *, char *, char *, int *); /*proto*/\nstatic CYTHON_INLINE int __pyx_f_5numpy_import_array(void); /*proto*/\n\n/* Module declarations from 'libc.stdlib' */\n\n/* Module declarations from 'pycocotools._mask' */\nstatic PyTypeObject *__pyx_ptype_11pycocotools_5_mask_RLEs = 0;\nstatic PyTypeObject *__pyx_ptype_11pycocotools_5_mask_Masks = 0;\nstatic __Pyx_TypeInfo __Pyx_TypeInfo_nn___pyx_t_5numpy_uint8_t = { \"uint8_t\", NULL, sizeof(__pyx_t_5numpy_uint8_t), { 0 }, 0, IS_UNSIGNED(__pyx_t_5numpy_uint8_t) ? 'U' : 'I', IS_UNSIGNED(__pyx_t_5numpy_uint8_t), 0 };\nstatic __Pyx_TypeInfo __Pyx_TypeInfo_nn___pyx_t_5numpy_double_t = { \"double_t\", NULL, sizeof(__pyx_t_5numpy_double_t), { 0 }, 0, 'R', 0, 0 };\nstatic __Pyx_TypeInfo __Pyx_TypeInfo_nn___pyx_t_5numpy_uint32_t = { \"uint32_t\", NULL, sizeof(__pyx_t_5numpy_uint32_t), { 0 }, 0, IS_UNSIGNED(__pyx_t_5numpy_uint32_t) ? 'U' : 'I', IS_UNSIGNED(__pyx_t_5numpy_uint32_t), 0 };\n#define __Pyx_MODULE_NAME \"pycocotools._mask\"\nextern int __pyx_module_is_main_pycocotools___mask;\nint __pyx_module_is_main_pycocotools___mask = 0;\n\n/* Implementation of 'pycocotools._mask' */\nstatic PyObject *__pyx_builtin_range;\nstatic PyObject *__pyx_builtin_AttributeError;\nstatic PyObject *__pyx_builtin_TypeError;\nstatic PyObject *__pyx_builtin_enumerate;\nstatic PyObject *__pyx_builtin_ValueError;\nstatic PyObject *__pyx_builtin_RuntimeError;\nstatic PyObject *__pyx_builtin_ImportError;\nstatic const char __pyx_k_F[] = \"F\";\nstatic const char __pyx_k_N[] = \"N\";\nstatic const char __pyx_k_R[] = \"R\";\nstatic const char __pyx_k_a[] = \"_a\";\nstatic const char __pyx_k_h[] = \"h\";\nstatic const char __pyx_k_i[] = \"i\";\nstatic const char __pyx_k_j[] = \"j\";\nstatic const char __pyx_k_m[] = \"m\";\nstatic const char __pyx_k_n[] = \"n\";\nstatic const char __pyx_k_p[] = \"p\";\nstatic const char __pyx_k_w[] = \"w\";\nstatic const char __pyx_k_Rs[] = \"Rs\";\nstatic const char __pyx_k_bb[] = \"bb\";\nstatic const char __pyx_k_dt[] = \"dt\";\nstatic const char __pyx_k_gt[] = \"gt\";\nstatic const char __pyx_k_np[] = \"np\";\nstatic const char __pyx_k_a_2[] = \"a\";\nstatic const char __pyx_k_all[] = \"all\";\nstatic const char __pyx_k_iou[] = \"_iou\";\nstatic const char __pyx_k_len[] = \"_len\";\nstatic const char __pyx_k_obj[] = \"obj\";\nstatic const char __pyx_k_RLEs[] = \"RLEs\";\nstatic const char __pyx_k_area[] = \"area\";\nstatic const char __pyx_k_bb_2[] = \"_bb\";\nstatic const char __pyx_k_cnts[] = \"cnts\";\nstatic const char __pyx_k_data[] = \"data\";\nstatic const char __pyx_k_main[] = \"__main__\";\nstatic const char __pyx_k_mask[] = \"mask\";\nstatic const char __pyx_k_name[] = \"__name__\";\nstatic const char __pyx_k_objs[] = \"objs\";\nstatic const char __pyx_k_poly[] = \"poly\";\nstatic const char __pyx_k_size[] = \"size\";\nstatic const char __pyx_k_test[] = \"__test__\";\nstatic const char __pyx_k_Masks[] = \"Masks\";\nstatic const char __pyx_k_array[] = \"array\";\nstatic const char __pyx_k_bbIou[] = \"_bbIou\";\nstatic const char __pyx_k_dtype[] = \"dtype\";\nstatic const char __pyx_k_iou_2[] = \"iou\";\nstatic const char __pyx_k_isbox[] = \"isbox\";\nstatic const char __pyx_k_isrle[] = \"isrle\";\nstatic const char __pyx_k_masks[] = \"masks\";\nstatic const char __pyx_k_merge[] = \"merge\";\nstatic const char __pyx_k_numpy[] = \"numpy\";\nstatic const char __pyx_k_order[] = \"order\";\nstatic const char __pyx_k_pyobj[] = \"pyobj\";\nstatic const char __pyx_k_range[] = \"range\";\nstatic const char __pyx_k_shape[] = \"shape\";\nstatic const char __pyx_k_uint8[] = \"uint8\";\nstatic const char __pyx_k_zeros[] = \"zeros\";\nstatic const char __pyx_k_astype[] = \"astype\";\nstatic const char __pyx_k_author[] = \"__author__\";\nstatic const char __pyx_k_counts[] = \"counts\";\nstatic const char __pyx_k_decode[] = \"decode\";\nstatic const char __pyx_k_double[] = \"double\";\nstatic const char __pyx_k_encode[] = \"encode\";\nstatic const char __pyx_k_frBbox[] = \"frBbox\";\nstatic const char __pyx_k_frPoly[] = \"frPoly\";\nstatic const char __pyx_k_import[] = \"__import__\";\nstatic const char __pyx_k_iouFun[] = \"_iouFun\";\nstatic const char __pyx_k_reduce[] = \"__reduce__\";\nstatic const char __pyx_k_rleIou[] = \"_rleIou\";\nstatic const char __pyx_k_toBbox[] = \"toBbox\";\nstatic const char __pyx_k_ucRles[] = \"ucRles\";\nstatic const char __pyx_k_uint32[] = \"uint32\";\nstatic const char __pyx_k_iscrowd[] = \"iscrowd\";\nstatic const char __pyx_k_np_poly[] = \"np_poly\";\nstatic const char __pyx_k_preproc[] = \"_preproc\";\nstatic const char __pyx_k_reshape[] = \"reshape\";\nstatic const char __pyx_k_rleObjs[] = \"rleObjs\";\nstatic const char __pyx_k_tsungyi[] = \"tsungyi\";\nstatic const char __pyx_k_c_string[] = \"c_string\";\nstatic const char __pyx_k_frString[] = \"_frString\";\nstatic const char __pyx_k_getstate[] = \"__getstate__\";\nstatic const char __pyx_k_setstate[] = \"__setstate__\";\nstatic const char __pyx_k_toString[] = \"_toString\";\nstatic const char __pyx_k_TypeError[] = \"TypeError\";\nstatic const char __pyx_k_enumerate[] = \"enumerate\";\nstatic const char __pyx_k_intersect[] = \"intersect\";\nstatic const char __pyx_k_py_string[] = \"py_string\";\nstatic const char __pyx_k_pyiscrowd[] = \"pyiscrowd\";\nstatic const char __pyx_k_reduce_ex[] = \"__reduce_ex__\";\nstatic const char __pyx_k_ValueError[] = \"ValueError\";\nstatic const char __pyx_k_ImportError[] = \"ImportError\";\nstatic const char __pyx_k_frPyObjects[] = \"frPyObjects\";\nstatic const char __pyx_k_RuntimeError[] = \"RuntimeError\";\nstatic const char __pyx_k_reduce_cython[] = \"__reduce_cython__\";\nstatic const char __pyx_k_AttributeError[] = \"AttributeError\";\nstatic const char __pyx_k_iou_locals__len[] = \"iou.._len\";\nstatic const char __pyx_k_setstate_cython[] = \"__setstate_cython__\";\nstatic const char __pyx_k_frUncompressedRLE[] = \"frUncompressedRLE\";\nstatic const char __pyx_k_iou_locals__bbIou[] = \"iou.._bbIou\";\nstatic const char __pyx_k_pycocotools__mask[] = \"pycocotools._mask\";\nstatic const char __pyx_k_cline_in_traceback[] = \"cline_in_traceback\";\nstatic const char __pyx_k_iou_locals__rleIou[] = \"iou.._rleIou\";\nstatic const char __pyx_k_iou_locals__preproc[] = \"iou.._preproc\";\nstatic const char __pyx_k_pycocotools__mask_pyx[] = \"pycocotools/_mask.pyx\";\nstatic const char __pyx_k_input_data_type_not_allowed[] = \"input data type not allowed.\";\nstatic const char __pyx_k_input_type_is_not_supported[] = \"input type is not supported.\";\nstatic const char __pyx_k_ndarray_is_not_C_contiguous[] = \"ndarray is not C contiguous\";\nstatic const char __pyx_k_numpy_core_multiarray_failed_to[] = \"numpy.core.multiarray failed to import\";\nstatic const char __pyx_k_numpy_ndarray_input_is_only_for[] = \"numpy ndarray input is only for *bounding boxes* and should have Nx4 dimension\";\nstatic const char __pyx_k_unknown_dtype_code_in_numpy_pxd[] = \"unknown dtype code in numpy.pxd (%d)\";\nstatic const char __pyx_k_unrecognized_type_The_following[] = \"unrecognized type. The following type: RLEs (rle), np.ndarray (box), and list (box) are supported.\";\nstatic const char __pyx_k_Format_string_allocated_too_shor[] = \"Format string allocated too short, see comment in numpy.pxd\";\nstatic const char __pyx_k_Non_native_byte_order_not_suppor[] = \"Non-native byte order not supported\";\nstatic const char __pyx_k_The_dt_and_gt_should_have_the_sa[] = \"The dt and gt should have the same data type, either RLEs, list or np.ndarray\";\nstatic const char __pyx_k_list_input_can_be_bounding_box_N[] = \"list input can be bounding box (Nx4) or RLEs ([RLE])\";\nstatic const char __pyx_k_ndarray_is_not_Fortran_contiguou[] = \"ndarray is not Fortran contiguous\";\nstatic const char __pyx_k_no_default___reduce___due_to_non[] = \"no default __reduce__ due to non-trivial __cinit__\";\nstatic const char __pyx_k_numpy_core_umath_failed_to_impor[] = \"numpy.core.umath failed to import\";\nstatic const char __pyx_k_Format_string_allocated_too_shor_2[] = \"Format string allocated too short.\";\nstatic PyObject *__pyx_n_s_AttributeError;\nstatic PyObject *__pyx_n_s_F;\nstatic PyObject *__pyx_kp_u_Format_string_allocated_too_shor;\nstatic PyObject *__pyx_kp_u_Format_string_allocated_too_shor_2;\nstatic PyObject *__pyx_n_s_ImportError;\nstatic PyObject *__pyx_n_s_Masks;\nstatic PyObject *__pyx_n_s_N;\nstatic PyObject *__pyx_kp_u_Non_native_byte_order_not_suppor;\nstatic PyObject *__pyx_n_s_R;\nstatic PyObject *__pyx_n_s_RLEs;\nstatic PyObject *__pyx_n_s_Rs;\nstatic PyObject *__pyx_n_s_RuntimeError;\nstatic PyObject *__pyx_kp_s_The_dt_and_gt_should_have_the_sa;\nstatic PyObject *__pyx_n_s_TypeError;\nstatic PyObject *__pyx_n_s_ValueError;\nstatic PyObject *__pyx_n_s_a;\nstatic PyObject *__pyx_n_s_a_2;\nstatic PyObject *__pyx_n_s_all;\nstatic PyObject *__pyx_n_s_area;\nstatic PyObject *__pyx_n_s_array;\nstatic PyObject *__pyx_n_s_astype;\nstatic PyObject *__pyx_n_s_author;\nstatic PyObject *__pyx_n_s_bb;\nstatic PyObject *__pyx_n_s_bbIou;\nstatic PyObject *__pyx_n_s_bb_2;\nstatic PyObject *__pyx_n_s_c_string;\nstatic PyObject *__pyx_n_s_cline_in_traceback;\nstatic PyObject *__pyx_n_s_cnts;\nstatic PyObject *__pyx_n_s_counts;\nstatic PyObject *__pyx_n_s_data;\nstatic PyObject *__pyx_n_s_decode;\nstatic PyObject *__pyx_n_s_double;\nstatic PyObject *__pyx_n_s_dt;\nstatic PyObject *__pyx_n_s_dtype;\nstatic PyObject *__pyx_n_s_encode;\nstatic PyObject *__pyx_n_s_enumerate;\nstatic PyObject *__pyx_n_s_frBbox;\nstatic PyObject *__pyx_n_s_frPoly;\nstatic PyObject *__pyx_n_s_frPyObjects;\nstatic PyObject *__pyx_n_s_frString;\nstatic PyObject *__pyx_n_s_frUncompressedRLE;\nstatic PyObject *__pyx_n_s_getstate;\nstatic PyObject *__pyx_n_s_gt;\nstatic PyObject *__pyx_n_s_h;\nstatic PyObject *__pyx_n_s_i;\nstatic PyObject *__pyx_n_s_import;\nstatic PyObject *__pyx_kp_s_input_data_type_not_allowed;\nstatic PyObject *__pyx_kp_s_input_type_is_not_supported;\nstatic PyObject *__pyx_n_s_intersect;\nstatic PyObject *__pyx_n_s_iou;\nstatic PyObject *__pyx_n_s_iouFun;\nstatic PyObject *__pyx_n_s_iou_2;\nstatic PyObject *__pyx_n_s_iou_locals__bbIou;\nstatic PyObject *__pyx_n_s_iou_locals__len;\nstatic PyObject *__pyx_n_s_iou_locals__preproc;\nstatic PyObject *__pyx_n_s_iou_locals__rleIou;\nstatic PyObject *__pyx_n_s_isbox;\nstatic PyObject *__pyx_n_s_iscrowd;\nstatic PyObject *__pyx_n_s_isrle;\nstatic PyObject *__pyx_n_s_j;\nstatic PyObject *__pyx_n_s_len;\nstatic PyObject *__pyx_kp_s_list_input_can_be_bounding_box_N;\nstatic PyObject *__pyx_n_s_m;\nstatic PyObject *__pyx_n_s_main;\nstatic PyObject *__pyx_n_s_mask;\nstatic PyObject *__pyx_n_s_masks;\nstatic PyObject *__pyx_n_s_merge;\nstatic PyObject *__pyx_n_s_n;\nstatic PyObject *__pyx_n_s_name;\nstatic PyObject *__pyx_kp_u_ndarray_is_not_C_contiguous;\nstatic PyObject *__pyx_kp_u_ndarray_is_not_Fortran_contiguou;\nstatic PyObject *__pyx_kp_s_no_default___reduce___due_to_non;\nstatic PyObject *__pyx_n_s_np;\nstatic PyObject *__pyx_n_s_np_poly;\nstatic PyObject *__pyx_n_s_numpy;\nstatic PyObject *__pyx_kp_s_numpy_core_multiarray_failed_to;\nstatic PyObject *__pyx_kp_s_numpy_core_umath_failed_to_impor;\nstatic PyObject *__pyx_kp_s_numpy_ndarray_input_is_only_for;\nstatic PyObject *__pyx_n_s_obj;\nstatic PyObject *__pyx_n_s_objs;\nstatic PyObject *__pyx_n_s_order;\nstatic PyObject *__pyx_n_s_p;\nstatic PyObject *__pyx_n_s_poly;\nstatic PyObject *__pyx_n_s_preproc;\nstatic PyObject *__pyx_n_s_py_string;\nstatic PyObject *__pyx_n_s_pycocotools__mask;\nstatic PyObject *__pyx_kp_s_pycocotools__mask_pyx;\nstatic PyObject *__pyx_n_s_pyiscrowd;\nstatic PyObject *__pyx_n_s_pyobj;\nstatic PyObject *__pyx_n_s_range;\nstatic PyObject *__pyx_n_s_reduce;\nstatic PyObject *__pyx_n_s_reduce_cython;\nstatic PyObject *__pyx_n_s_reduce_ex;\nstatic PyObject *__pyx_n_s_reshape;\nstatic PyObject *__pyx_n_s_rleIou;\nstatic PyObject *__pyx_n_s_rleObjs;\nstatic PyObject *__pyx_n_s_setstate;\nstatic PyObject *__pyx_n_s_setstate_cython;\nstatic PyObject *__pyx_n_s_shape;\nstatic PyObject *__pyx_n_s_size;\nstatic PyObject *__pyx_n_s_test;\nstatic PyObject *__pyx_n_s_toBbox;\nstatic PyObject *__pyx_n_s_toString;\nstatic PyObject *__pyx_n_s_tsungyi;\nstatic PyObject *__pyx_n_s_ucRles;\nstatic PyObject *__pyx_n_s_uint32;\nstatic PyObject *__pyx_n_s_uint8;\nstatic PyObject *__pyx_kp_u_unknown_dtype_code_in_numpy_pxd;\nstatic PyObject *__pyx_kp_s_unrecognized_type_The_following;\nstatic PyObject *__pyx_n_s_w;\nstatic PyObject *__pyx_n_s_zeros;\nstatic int __pyx_pf_11pycocotools_5_mask_4RLEs___cinit__(struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_self, siz __pyx_v_n); /* proto */\nstatic void __pyx_pf_11pycocotools_5_mask_4RLEs_2__dealloc__(struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_self); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_4RLEs_4__getattr__(struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_self, PyObject *__pyx_v_key); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_4RLEs_6__reduce_cython__(CYTHON_UNUSED struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_self); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_4RLEs_8__setstate_cython__(CYTHON_UNUSED struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_self, CYTHON_UNUSED PyObject *__pyx_v___pyx_state); /* proto */\nstatic int __pyx_pf_11pycocotools_5_mask_5Masks___cinit__(struct __pyx_obj_11pycocotools_5_mask_Masks *__pyx_v_self, PyObject *__pyx_v_h, PyObject *__pyx_v_w, PyObject *__pyx_v_n); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_5Masks_2__array__(struct __pyx_obj_11pycocotools_5_mask_Masks *__pyx_v_self); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_5Masks_4__reduce_cython__(CYTHON_UNUSED struct __pyx_obj_11pycocotools_5_mask_Masks *__pyx_v_self); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_5Masks_6__setstate_cython__(CYTHON_UNUSED struct __pyx_obj_11pycocotools_5_mask_Masks *__pyx_v_self, CYTHON_UNUSED PyObject *__pyx_v___pyx_state); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask__toString(CYTHON_UNUSED PyObject *__pyx_self, struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_Rs); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_2_frString(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_rleObjs); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_4encode(CYTHON_UNUSED PyObject *__pyx_self, PyArrayObject *__pyx_v_mask); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_6decode(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_rleObjs); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_8merge(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_rleObjs, int __pyx_v_intersect); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_10area(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_rleObjs); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_3iou__preproc(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_objs); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_3iou_2_rleIou(CYTHON_UNUSED PyObject *__pyx_self, struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_dt, struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_gt, PyArrayObject *__pyx_v_iscrowd, siz __pyx_v_m, siz __pyx_v_n, PyArrayObject *__pyx_v__iou); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_3iou_4_bbIou(CYTHON_UNUSED PyObject *__pyx_self, PyArrayObject *__pyx_v_dt, PyArrayObject *__pyx_v_gt, PyArrayObject *__pyx_v_iscrowd, siz __pyx_v_m, siz __pyx_v_n, PyArrayObject *__pyx_v__iou); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_3iou_6_len(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_obj); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_12iou(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_dt, PyObject *__pyx_v_gt, PyObject *__pyx_v_pyiscrowd); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_14toBbox(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_rleObjs); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_16frBbox(CYTHON_UNUSED PyObject *__pyx_self, PyArrayObject *__pyx_v_bb, siz __pyx_v_h, siz __pyx_v_w); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_18frPoly(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_poly, siz __pyx_v_h, siz __pyx_v_w); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_20frUncompressedRLE(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_ucRles, CYTHON_UNUSED siz __pyx_v_h, CYTHON_UNUSED siz __pyx_v_w); /* proto */\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_22frPyObjects(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_pyobj, siz __pyx_v_h, PyObject *__pyx_v_w); /* proto */\nstatic int __pyx_pf_5numpy_7ndarray___getbuffer__(PyArrayObject *__pyx_v_self, Py_buffer *__pyx_v_info, int __pyx_v_flags); /* proto */\nstatic void __pyx_pf_5numpy_7ndarray_2__releasebuffer__(PyArrayObject *__pyx_v_self, Py_buffer *__pyx_v_info); /* proto */\nstatic PyObject *__pyx_tp_new_11pycocotools_5_mask_RLEs(PyTypeObject *t, PyObject *a, PyObject *k); /*proto*/\nstatic PyObject *__pyx_tp_new_11pycocotools_5_mask_Masks(PyTypeObject *t, PyObject *a, PyObject *k); /*proto*/\nstatic PyObject *__pyx_int_0;\nstatic PyObject *__pyx_int_1;\nstatic PyObject *__pyx_int_4;\nstatic PyObject *__pyx_tuple_;\nstatic PyObject *__pyx_tuple__2;\nstatic PyObject *__pyx_tuple__3;\nstatic PyObject *__pyx_tuple__4;\nstatic PyObject *__pyx_tuple__5;\nstatic PyObject *__pyx_tuple__6;\nstatic PyObject *__pyx_tuple__7;\nstatic PyObject *__pyx_tuple__8;\nstatic PyObject *__pyx_tuple__9;\nstatic PyObject *__pyx_tuple__11;\nstatic PyObject *__pyx_tuple__13;\nstatic PyObject *__pyx_tuple__15;\nstatic PyObject *__pyx_tuple__17;\nstatic PyObject *__pyx_tuple__18;\nstatic PyObject *__pyx_tuple__19;\nstatic PyObject *__pyx_tuple__20;\nstatic PyObject *__pyx_tuple__21;\nstatic PyObject *__pyx_tuple__22;\nstatic PyObject *__pyx_tuple__23;\nstatic PyObject *__pyx_tuple__24;\nstatic PyObject *__pyx_tuple__25;\nstatic PyObject *__pyx_tuple__26;\nstatic PyObject *__pyx_tuple__27;\nstatic PyObject *__pyx_tuple__29;\nstatic PyObject *__pyx_tuple__31;\nstatic PyObject *__pyx_tuple__33;\nstatic PyObject *__pyx_tuple__35;\nstatic PyObject *__pyx_tuple__37;\nstatic PyObject *__pyx_tuple__39;\nstatic PyObject *__pyx_tuple__41;\nstatic PyObject *__pyx_tuple__43;\nstatic PyObject *__pyx_tuple__45;\nstatic PyObject *__pyx_tuple__47;\nstatic PyObject *__pyx_tuple__49;\nstatic PyObject *__pyx_codeobj__10;\nstatic PyObject *__pyx_codeobj__12;\nstatic PyObject *__pyx_codeobj__14;\nstatic PyObject *__pyx_codeobj__16;\nstatic PyObject *__pyx_codeobj__28;\nstatic PyObject *__pyx_codeobj__30;\nstatic PyObject *__pyx_codeobj__32;\nstatic PyObject *__pyx_codeobj__34;\nstatic PyObject *__pyx_codeobj__36;\nstatic PyObject *__pyx_codeobj__38;\nstatic PyObject *__pyx_codeobj__40;\nstatic PyObject *__pyx_codeobj__42;\nstatic PyObject *__pyx_codeobj__44;\nstatic PyObject *__pyx_codeobj__46;\nstatic PyObject *__pyx_codeobj__48;\nstatic PyObject *__pyx_codeobj__50;\n/* Late includes */\n\n/* \"pycocotools/_mask.pyx\":57\n * cdef siz _n\n * \n * def __cinit__(self, siz n =0): # <<<<<<<<<<<<<<\n * rlesInit(&self._R, n)\n * self._n = n\n */\n\n/* Python wrapper */\nstatic int __pyx_pw_11pycocotools_5_mask_4RLEs_1__cinit__(PyObject *__pyx_v_self, PyObject *__pyx_args, PyObject *__pyx_kwds); /*proto*/\nstatic int __pyx_pw_11pycocotools_5_mask_4RLEs_1__cinit__(PyObject *__pyx_v_self, PyObject *__pyx_args, PyObject *__pyx_kwds) {\n siz __pyx_v_n;\n int __pyx_r;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"__cinit__ (wrapper)\", 0);\n {\n static PyObject **__pyx_pyargnames[] = {&__pyx_n_s_n,0};\n PyObject* values[1] = {0};\n if (unlikely(__pyx_kwds)) {\n Py_ssize_t kw_args;\n const Py_ssize_t pos_args = PyTuple_GET_SIZE(__pyx_args);\n switch (pos_args) {\n case 1: values[0] = PyTuple_GET_ITEM(__pyx_args, 0);\n CYTHON_FALLTHROUGH;\n case 0: break;\n default: goto __pyx_L5_argtuple_error;\n }\n kw_args = PyDict_Size(__pyx_kwds);\n switch (pos_args) {\n case 0:\n if (kw_args > 0) {\n PyObject* value = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_n);\n if (value) { values[0] = value; kw_args--; }\n }\n }\n if (unlikely(kw_args > 0)) {\n if (unlikely(__Pyx_ParseOptionalKeywords(__pyx_kwds, __pyx_pyargnames, 0, values, pos_args, \"__cinit__\") < 0)) __PYX_ERR(0, 57, __pyx_L3_error)\n }\n } else {\n switch (PyTuple_GET_SIZE(__pyx_args)) {\n case 1: values[0] = PyTuple_GET_ITEM(__pyx_args, 0);\n CYTHON_FALLTHROUGH;\n case 0: break;\n default: goto __pyx_L5_argtuple_error;\n }\n }\n if (values[0]) {\n __pyx_v_n = __Pyx_PyInt_As_siz(values[0]); if (unlikely((__pyx_v_n == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 57, __pyx_L3_error)\n } else {\n __pyx_v_n = ((siz)0);\n }\n }\n goto __pyx_L4_argument_unpacking_done;\n __pyx_L5_argtuple_error:;\n __Pyx_RaiseArgtupleInvalid(\"__cinit__\", 0, 0, 1, PyTuple_GET_SIZE(__pyx_args)); __PYX_ERR(0, 57, __pyx_L3_error)\n __pyx_L3_error:;\n __Pyx_AddTraceback(\"pycocotools._mask.RLEs.__cinit__\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __Pyx_RefNannyFinishContext();\n return -1;\n __pyx_L4_argument_unpacking_done:;\n __pyx_r = __pyx_pf_11pycocotools_5_mask_4RLEs___cinit__(((struct __pyx_obj_11pycocotools_5_mask_RLEs *)__pyx_v_self), __pyx_v_n);\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic int __pyx_pf_11pycocotools_5_mask_4RLEs___cinit__(struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_self, siz __pyx_v_n) {\n int __pyx_r;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"__cinit__\", 0);\n\n /* \"pycocotools/_mask.pyx\":58\n * \n * def __cinit__(self, siz n =0):\n * rlesInit(&self._R, n) # <<<<<<<<<<<<<<\n * self._n = n\n * \n */\n rlesInit((&__pyx_v_self->_R), __pyx_v_n);\n\n /* \"pycocotools/_mask.pyx\":59\n * def __cinit__(self, siz n =0):\n * rlesInit(&self._R, n)\n * self._n = n # <<<<<<<<<<<<<<\n * \n * # free the RLE array here\n */\n __pyx_v_self->_n = __pyx_v_n;\n\n /* \"pycocotools/_mask.pyx\":57\n * cdef siz _n\n * \n * def __cinit__(self, siz n =0): # <<<<<<<<<<<<<<\n * rlesInit(&self._R, n)\n * self._n = n\n */\n\n /* function exit code */\n __pyx_r = 0;\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"pycocotools/_mask.pyx\":62\n * \n * # free the RLE array here\n * def __dealloc__(self): # <<<<<<<<<<<<<<\n * if self._R is not NULL:\n * for i in range(self._n):\n */\n\n/* Python wrapper */\nstatic void __pyx_pw_11pycocotools_5_mask_4RLEs_3__dealloc__(PyObject *__pyx_v_self); /*proto*/\nstatic void __pyx_pw_11pycocotools_5_mask_4RLEs_3__dealloc__(PyObject *__pyx_v_self) {\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"__dealloc__ (wrapper)\", 0);\n __pyx_pf_11pycocotools_5_mask_4RLEs_2__dealloc__(((struct __pyx_obj_11pycocotools_5_mask_RLEs *)__pyx_v_self));\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n}\n\nstatic void __pyx_pf_11pycocotools_5_mask_4RLEs_2__dealloc__(struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_self) {\n siz __pyx_v_i;\n __Pyx_RefNannyDeclarations\n int __pyx_t_1;\n siz __pyx_t_2;\n siz __pyx_t_3;\n siz __pyx_t_4;\n __Pyx_RefNannySetupContext(\"__dealloc__\", 0);\n\n /* \"pycocotools/_mask.pyx\":63\n * # free the RLE array here\n * def __dealloc__(self):\n * if self._R is not NULL: # <<<<<<<<<<<<<<\n * for i in range(self._n):\n * free(self._R[i].cnts)\n */\n __pyx_t_1 = ((__pyx_v_self->_R != NULL) != 0);\n if (__pyx_t_1) {\n\n /* \"pycocotools/_mask.pyx\":64\n * def __dealloc__(self):\n * if self._R is not NULL:\n * for i in range(self._n): # <<<<<<<<<<<<<<\n * free(self._R[i].cnts)\n * free(self._R)\n */\n __pyx_t_2 = __pyx_v_self->_n;\n __pyx_t_3 = __pyx_t_2;\n for (__pyx_t_4 = 0; __pyx_t_4 < __pyx_t_3; __pyx_t_4+=1) {\n __pyx_v_i = __pyx_t_4;\n\n /* \"pycocotools/_mask.pyx\":65\n * if self._R is not NULL:\n * for i in range(self._n):\n * free(self._R[i].cnts) # <<<<<<<<<<<<<<\n * free(self._R)\n * def __getattr__(self, key):\n */\n free((__pyx_v_self->_R[__pyx_v_i]).cnts);\n }\n\n /* \"pycocotools/_mask.pyx\":66\n * for i in range(self._n):\n * free(self._R[i].cnts)\n * free(self._R) # <<<<<<<<<<<<<<\n * def __getattr__(self, key):\n * if key == 'n':\n */\n free(__pyx_v_self->_R);\n\n /* \"pycocotools/_mask.pyx\":63\n * # free the RLE array here\n * def __dealloc__(self):\n * if self._R is not NULL: # <<<<<<<<<<<<<<\n * for i in range(self._n):\n * free(self._R[i].cnts)\n */\n }\n\n /* \"pycocotools/_mask.pyx\":62\n * \n * # free the RLE array here\n * def __dealloc__(self): # <<<<<<<<<<<<<<\n * if self._R is not NULL:\n * for i in range(self._n):\n */\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n}\n\n/* \"pycocotools/_mask.pyx\":67\n * free(self._R[i].cnts)\n * free(self._R)\n * def __getattr__(self, key): # <<<<<<<<<<<<<<\n * if key == 'n':\n * return self._n\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_4RLEs_5__getattr__(PyObject *__pyx_v_self, PyObject *__pyx_v_key); /*proto*/\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_4RLEs_5__getattr__(PyObject *__pyx_v_self, PyObject *__pyx_v_key) {\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"__getattr__ (wrapper)\", 0);\n __pyx_r = __pyx_pf_11pycocotools_5_mask_4RLEs_4__getattr__(((struct __pyx_obj_11pycocotools_5_mask_RLEs *)__pyx_v_self), ((PyObject *)__pyx_v_key));\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_4RLEs_4__getattr__(struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_self, PyObject *__pyx_v_key) {\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n int __pyx_t_1;\n PyObject *__pyx_t_2 = NULL;\n __Pyx_RefNannySetupContext(\"__getattr__\", 0);\n\n /* \"pycocotools/_mask.pyx\":68\n * free(self._R)\n * def __getattr__(self, key):\n * if key == 'n': # <<<<<<<<<<<<<<\n * return self._n\n * raise AttributeError(key)\n */\n __pyx_t_1 = (__Pyx_PyString_Equals(__pyx_v_key, __pyx_n_s_n, Py_EQ)); if (unlikely(__pyx_t_1 < 0)) __PYX_ERR(0, 68, __pyx_L1_error)\n if (__pyx_t_1) {\n\n /* \"pycocotools/_mask.pyx\":69\n * def __getattr__(self, key):\n * if key == 'n':\n * return self._n # <<<<<<<<<<<<<<\n * raise AttributeError(key)\n * \n */\n __Pyx_XDECREF(__pyx_r);\n __pyx_t_2 = __Pyx_PyInt_From_siz(__pyx_v_self->_n); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 69, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __pyx_r = __pyx_t_2;\n __pyx_t_2 = 0;\n goto __pyx_L0;\n\n /* \"pycocotools/_mask.pyx\":68\n * free(self._R)\n * def __getattr__(self, key):\n * if key == 'n': # <<<<<<<<<<<<<<\n * return self._n\n * raise AttributeError(key)\n */\n }\n\n /* \"pycocotools/_mask.pyx\":70\n * if key == 'n':\n * return self._n\n * raise AttributeError(key) # <<<<<<<<<<<<<<\n * \n * # python class to wrap Mask array in C\n */\n __pyx_t_2 = __Pyx_PyObject_CallOneArg(__pyx_builtin_AttributeError, __pyx_v_key); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 70, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __Pyx_Raise(__pyx_t_2, 0, 0, 0);\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n __PYX_ERR(0, 70, __pyx_L1_error)\n\n /* \"pycocotools/_mask.pyx\":67\n * free(self._R[i].cnts)\n * free(self._R)\n * def __getattr__(self, key): # <<<<<<<<<<<<<<\n * if key == 'n':\n * return self._n\n */\n\n /* function exit code */\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_2);\n __Pyx_AddTraceback(\"pycocotools._mask.RLEs.__getattr__\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n __pyx_L0:;\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"(tree fragment)\":1\n * def __reduce_cython__(self): # <<<<<<<<<<<<<<\n * raise TypeError(\"no default __reduce__ due to non-trivial __cinit__\")\n * def __setstate_cython__(self, __pyx_state):\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_4RLEs_7__reduce_cython__(PyObject *__pyx_v_self, CYTHON_UNUSED PyObject *unused); /*proto*/\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_4RLEs_7__reduce_cython__(PyObject *__pyx_v_self, CYTHON_UNUSED PyObject *unused) {\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"__reduce_cython__ (wrapper)\", 0);\n __pyx_r = __pyx_pf_11pycocotools_5_mask_4RLEs_6__reduce_cython__(((struct __pyx_obj_11pycocotools_5_mask_RLEs *)__pyx_v_self));\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_4RLEs_6__reduce_cython__(CYTHON_UNUSED struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_self) {\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n PyObject *__pyx_t_1 = NULL;\n __Pyx_RefNannySetupContext(\"__reduce_cython__\", 0);\n\n /* \"(tree fragment)\":2\n * def __reduce_cython__(self):\n * raise TypeError(\"no default __reduce__ due to non-trivial __cinit__\") # <<<<<<<<<<<<<<\n * def __setstate_cython__(self, __pyx_state):\n * raise TypeError(\"no default __reduce__ due to non-trivial __cinit__\")\n */\n __pyx_t_1 = __Pyx_PyObject_Call(__pyx_builtin_TypeError, __pyx_tuple_, NULL); if (unlikely(!__pyx_t_1)) __PYX_ERR(1, 2, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_Raise(__pyx_t_1, 0, 0, 0);\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __PYX_ERR(1, 2, __pyx_L1_error)\n\n /* \"(tree fragment)\":1\n * def __reduce_cython__(self): # <<<<<<<<<<<<<<\n * raise TypeError(\"no default __reduce__ due to non-trivial __cinit__\")\n * def __setstate_cython__(self, __pyx_state):\n */\n\n /* function exit code */\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_1);\n __Pyx_AddTraceback(\"pycocotools._mask.RLEs.__reduce_cython__\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"(tree fragment)\":3\n * def __reduce_cython__(self):\n * raise TypeError(\"no default __reduce__ due to non-trivial __cinit__\")\n * def __setstate_cython__(self, __pyx_state): # <<<<<<<<<<<<<<\n * raise TypeError(\"no default __reduce__ due to non-trivial __cinit__\")\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_4RLEs_9__setstate_cython__(PyObject *__pyx_v_self, PyObject *__pyx_v___pyx_state); /*proto*/\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_4RLEs_9__setstate_cython__(PyObject *__pyx_v_self, PyObject *__pyx_v___pyx_state) {\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"__setstate_cython__ (wrapper)\", 0);\n __pyx_r = __pyx_pf_11pycocotools_5_mask_4RLEs_8__setstate_cython__(((struct __pyx_obj_11pycocotools_5_mask_RLEs *)__pyx_v_self), ((PyObject *)__pyx_v___pyx_state));\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_4RLEs_8__setstate_cython__(CYTHON_UNUSED struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_self, CYTHON_UNUSED PyObject *__pyx_v___pyx_state) {\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n PyObject *__pyx_t_1 = NULL;\n __Pyx_RefNannySetupContext(\"__setstate_cython__\", 0);\n\n /* \"(tree fragment)\":4\n * raise TypeError(\"no default __reduce__ due to non-trivial __cinit__\")\n * def __setstate_cython__(self, __pyx_state):\n * raise TypeError(\"no default __reduce__ due to non-trivial __cinit__\") # <<<<<<<<<<<<<<\n */\n __pyx_t_1 = __Pyx_PyObject_Call(__pyx_builtin_TypeError, __pyx_tuple__2, NULL); if (unlikely(!__pyx_t_1)) __PYX_ERR(1, 4, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_Raise(__pyx_t_1, 0, 0, 0);\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __PYX_ERR(1, 4, __pyx_L1_error)\n\n /* \"(tree fragment)\":3\n * def __reduce_cython__(self):\n * raise TypeError(\"no default __reduce__ due to non-trivial __cinit__\")\n * def __setstate_cython__(self, __pyx_state): # <<<<<<<<<<<<<<\n * raise TypeError(\"no default __reduce__ due to non-trivial __cinit__\")\n */\n\n /* function exit code */\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_1);\n __Pyx_AddTraceback(\"pycocotools._mask.RLEs.__setstate_cython__\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"pycocotools/_mask.pyx\":80\n * cdef siz _n\n * \n * def __cinit__(self, h, w, n): # <<<<<<<<<<<<<<\n * self._mask = malloc(h*w*n* sizeof(byte))\n * self._h = h\n */\n\n/* Python wrapper */\nstatic int __pyx_pw_11pycocotools_5_mask_5Masks_1__cinit__(PyObject *__pyx_v_self, PyObject *__pyx_args, PyObject *__pyx_kwds); /*proto*/\nstatic int __pyx_pw_11pycocotools_5_mask_5Masks_1__cinit__(PyObject *__pyx_v_self, PyObject *__pyx_args, PyObject *__pyx_kwds) {\n PyObject *__pyx_v_h = 0;\n PyObject *__pyx_v_w = 0;\n PyObject *__pyx_v_n = 0;\n int __pyx_r;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"__cinit__ (wrapper)\", 0);\n {\n static PyObject **__pyx_pyargnames[] = {&__pyx_n_s_h,&__pyx_n_s_w,&__pyx_n_s_n,0};\n PyObject* values[3] = {0,0,0};\n if (unlikely(__pyx_kwds)) {\n Py_ssize_t kw_args;\n const Py_ssize_t pos_args = PyTuple_GET_SIZE(__pyx_args);\n switch (pos_args) {\n case 3: values[2] = PyTuple_GET_ITEM(__pyx_args, 2);\n CYTHON_FALLTHROUGH;\n case 2: values[1] = PyTuple_GET_ITEM(__pyx_args, 1);\n CYTHON_FALLTHROUGH;\n case 1: values[0] = PyTuple_GET_ITEM(__pyx_args, 0);\n CYTHON_FALLTHROUGH;\n case 0: break;\n default: goto __pyx_L5_argtuple_error;\n }\n kw_args = PyDict_Size(__pyx_kwds);\n switch (pos_args) {\n case 0:\n if (likely((values[0] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_h)) != 0)) kw_args--;\n else goto __pyx_L5_argtuple_error;\n CYTHON_FALLTHROUGH;\n case 1:\n if (likely((values[1] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_w)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"__cinit__\", 1, 3, 3, 1); __PYX_ERR(0, 80, __pyx_L3_error)\n }\n CYTHON_FALLTHROUGH;\n case 2:\n if (likely((values[2] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_n)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"__cinit__\", 1, 3, 3, 2); __PYX_ERR(0, 80, __pyx_L3_error)\n }\n }\n if (unlikely(kw_args > 0)) {\n if (unlikely(__Pyx_ParseOptionalKeywords(__pyx_kwds, __pyx_pyargnames, 0, values, pos_args, \"__cinit__\") < 0)) __PYX_ERR(0, 80, __pyx_L3_error)\n }\n } else if (PyTuple_GET_SIZE(__pyx_args) != 3) {\n goto __pyx_L5_argtuple_error;\n } else {\n values[0] = PyTuple_GET_ITEM(__pyx_args, 0);\n values[1] = PyTuple_GET_ITEM(__pyx_args, 1);\n values[2] = PyTuple_GET_ITEM(__pyx_args, 2);\n }\n __pyx_v_h = values[0];\n __pyx_v_w = values[1];\n __pyx_v_n = values[2];\n }\n goto __pyx_L4_argument_unpacking_done;\n __pyx_L5_argtuple_error:;\n __Pyx_RaiseArgtupleInvalid(\"__cinit__\", 1, 3, 3, PyTuple_GET_SIZE(__pyx_args)); __PYX_ERR(0, 80, __pyx_L3_error)\n __pyx_L3_error:;\n __Pyx_AddTraceback(\"pycocotools._mask.Masks.__cinit__\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __Pyx_RefNannyFinishContext();\n return -1;\n __pyx_L4_argument_unpacking_done:;\n __pyx_r = __pyx_pf_11pycocotools_5_mask_5Masks___cinit__(((struct __pyx_obj_11pycocotools_5_mask_Masks *)__pyx_v_self), __pyx_v_h, __pyx_v_w, __pyx_v_n);\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic int __pyx_pf_11pycocotools_5_mask_5Masks___cinit__(struct __pyx_obj_11pycocotools_5_mask_Masks *__pyx_v_self, PyObject *__pyx_v_h, PyObject *__pyx_v_w, PyObject *__pyx_v_n) {\n int __pyx_r;\n __Pyx_RefNannyDeclarations\n PyObject *__pyx_t_1 = NULL;\n PyObject *__pyx_t_2 = NULL;\n PyObject *__pyx_t_3 = NULL;\n size_t __pyx_t_4;\n siz __pyx_t_5;\n __Pyx_RefNannySetupContext(\"__cinit__\", 0);\n\n /* \"pycocotools/_mask.pyx\":81\n * \n * def __cinit__(self, h, w, n):\n * self._mask = malloc(h*w*n* sizeof(byte)) # <<<<<<<<<<<<<<\n * self._h = h\n * self._w = w\n */\n __pyx_t_1 = PyNumber_Multiply(__pyx_v_h, __pyx_v_w); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 81, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_2 = PyNumber_Multiply(__pyx_t_1, __pyx_v_n); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 81, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __pyx_t_1 = __Pyx_PyInt_FromSize_t((sizeof(byte))); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 81, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_3 = PyNumber_Multiply(__pyx_t_2, __pyx_t_1); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 81, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __pyx_t_4 = __Pyx_PyInt_As_size_t(__pyx_t_3); if (unlikely((__pyx_t_4 == (size_t)-1) && PyErr_Occurred())) __PYX_ERR(0, 81, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __pyx_v_self->_mask = ((byte *)malloc(__pyx_t_4));\n\n /* \"pycocotools/_mask.pyx\":82\n * def __cinit__(self, h, w, n):\n * self._mask = malloc(h*w*n* sizeof(byte))\n * self._h = h # <<<<<<<<<<<<<<\n * self._w = w\n * self._n = n\n */\n __pyx_t_5 = __Pyx_PyInt_As_siz(__pyx_v_h); if (unlikely((__pyx_t_5 == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 82, __pyx_L1_error)\n __pyx_v_self->_h = __pyx_t_5;\n\n /* \"pycocotools/_mask.pyx\":83\n * self._mask = malloc(h*w*n* sizeof(byte))\n * self._h = h\n * self._w = w # <<<<<<<<<<<<<<\n * self._n = n\n * # def __dealloc__(self):\n */\n __pyx_t_5 = __Pyx_PyInt_As_siz(__pyx_v_w); if (unlikely((__pyx_t_5 == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 83, __pyx_L1_error)\n __pyx_v_self->_w = __pyx_t_5;\n\n /* \"pycocotools/_mask.pyx\":84\n * self._h = h\n * self._w = w\n * self._n = n # <<<<<<<<<<<<<<\n * # def __dealloc__(self):\n * # the memory management of _mask has been passed to np.ndarray\n */\n __pyx_t_5 = __Pyx_PyInt_As_siz(__pyx_v_n); if (unlikely((__pyx_t_5 == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 84, __pyx_L1_error)\n __pyx_v_self->_n = __pyx_t_5;\n\n /* \"pycocotools/_mask.pyx\":80\n * cdef siz _n\n * \n * def __cinit__(self, h, w, n): # <<<<<<<<<<<<<<\n * self._mask = malloc(h*w*n* sizeof(byte))\n * self._h = h\n */\n\n /* function exit code */\n __pyx_r = 0;\n goto __pyx_L0;\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_1);\n __Pyx_XDECREF(__pyx_t_2);\n __Pyx_XDECREF(__pyx_t_3);\n __Pyx_AddTraceback(\"pycocotools._mask.Masks.__cinit__\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = -1;\n __pyx_L0:;\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"pycocotools/_mask.pyx\":90\n * \n * # called when passing into np.array() and return an np.ndarray in column-major order\n * def __array__(self): # <<<<<<<<<<<<<<\n * cdef np.npy_intp shape[1]\n * shape[0] = self._h*self._w*self._n\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_5Masks_3__array__(PyObject *__pyx_v_self, CYTHON_UNUSED PyObject *unused); /*proto*/\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_5Masks_3__array__(PyObject *__pyx_v_self, CYTHON_UNUSED PyObject *unused) {\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"__array__ (wrapper)\", 0);\n __pyx_r = __pyx_pf_11pycocotools_5_mask_5Masks_2__array__(((struct __pyx_obj_11pycocotools_5_mask_Masks *)__pyx_v_self));\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_5Masks_2__array__(struct __pyx_obj_11pycocotools_5_mask_Masks *__pyx_v_self) {\n npy_intp __pyx_v_shape[1];\n PyObject *__pyx_v_ndarray = NULL;\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n PyObject *__pyx_t_1 = NULL;\n PyObject *__pyx_t_2 = NULL;\n PyObject *__pyx_t_3 = NULL;\n PyObject *__pyx_t_4 = NULL;\n PyObject *__pyx_t_5 = NULL;\n __Pyx_RefNannySetupContext(\"__array__\", 0);\n\n /* \"pycocotools/_mask.pyx\":92\n * def __array__(self):\n * cdef np.npy_intp shape[1]\n * shape[0] = self._h*self._w*self._n # <<<<<<<<<<<<<<\n * # Create a 1D array, and reshape it to fortran/Matlab column-major array\n * ndarray = np.PyArray_SimpleNewFromData(1, shape, np.NPY_UINT8, self._mask).reshape((self._h, self._w, self._n), order='F')\n */\n (__pyx_v_shape[0]) = ((((npy_intp)__pyx_v_self->_h) * __pyx_v_self->_w) * __pyx_v_self->_n);\n\n /* \"pycocotools/_mask.pyx\":94\n * shape[0] = self._h*self._w*self._n\n * # Create a 1D array, and reshape it to fortran/Matlab column-major array\n * ndarray = np.PyArray_SimpleNewFromData(1, shape, np.NPY_UINT8, self._mask).reshape((self._h, self._w, self._n), order='F') # <<<<<<<<<<<<<<\n * # The _mask allocated by Masks is now handled by ndarray\n * PyArray_ENABLEFLAGS(ndarray, np.NPY_OWNDATA)\n */\n __pyx_t_1 = PyArray_SimpleNewFromData(1, __pyx_v_shape, NPY_UINT8, __pyx_v_self->_mask); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 94, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_2 = __Pyx_PyObject_GetAttrStr(__pyx_t_1, __pyx_n_s_reshape); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 94, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __pyx_t_1 = __Pyx_PyInt_From_siz(__pyx_v_self->_h); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 94, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_3 = __Pyx_PyInt_From_siz(__pyx_v_self->_w); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 94, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __pyx_t_4 = __Pyx_PyInt_From_siz(__pyx_v_self->_n); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 94, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __pyx_t_5 = PyTuple_New(3); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 94, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_GIVEREF(__pyx_t_1);\n PyTuple_SET_ITEM(__pyx_t_5, 0, __pyx_t_1);\n __Pyx_GIVEREF(__pyx_t_3);\n PyTuple_SET_ITEM(__pyx_t_5, 1, __pyx_t_3);\n __Pyx_GIVEREF(__pyx_t_4);\n PyTuple_SET_ITEM(__pyx_t_5, 2, __pyx_t_4);\n __pyx_t_1 = 0;\n __pyx_t_3 = 0;\n __pyx_t_4 = 0;\n __pyx_t_4 = PyTuple_New(1); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 94, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __Pyx_GIVEREF(__pyx_t_5);\n PyTuple_SET_ITEM(__pyx_t_4, 0, __pyx_t_5);\n __pyx_t_5 = 0;\n __pyx_t_5 = __Pyx_PyDict_NewPresized(1); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 94, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n if (PyDict_SetItem(__pyx_t_5, __pyx_n_s_order, __pyx_n_s_F) < 0) __PYX_ERR(0, 94, __pyx_L1_error)\n __pyx_t_3 = __Pyx_PyObject_Call(__pyx_t_2, __pyx_t_4, __pyx_t_5); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 94, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __pyx_v_ndarray = __pyx_t_3;\n __pyx_t_3 = 0;\n\n /* \"pycocotools/_mask.pyx\":96\n * ndarray = np.PyArray_SimpleNewFromData(1, shape, np.NPY_UINT8, self._mask).reshape((self._h, self._w, self._n), order='F')\n * # The _mask allocated by Masks is now handled by ndarray\n * PyArray_ENABLEFLAGS(ndarray, np.NPY_OWNDATA) # <<<<<<<<<<<<<<\n * return ndarray\n * \n */\n if (!(likely(((__pyx_v_ndarray) == Py_None) || likely(__Pyx_TypeTest(__pyx_v_ndarray, __pyx_ptype_5numpy_ndarray))))) __PYX_ERR(0, 96, __pyx_L1_error)\n PyArray_ENABLEFLAGS(((PyArrayObject *)__pyx_v_ndarray), NPY_OWNDATA);\n\n /* \"pycocotools/_mask.pyx\":97\n * # The _mask allocated by Masks is now handled by ndarray\n * PyArray_ENABLEFLAGS(ndarray, np.NPY_OWNDATA)\n * return ndarray # <<<<<<<<<<<<<<\n * \n * # internal conversion from Python RLEs object to compressed RLE format\n */\n __Pyx_XDECREF(__pyx_r);\n __Pyx_INCREF(__pyx_v_ndarray);\n __pyx_r = __pyx_v_ndarray;\n goto __pyx_L0;\n\n /* \"pycocotools/_mask.pyx\":90\n * \n * # called when passing into np.array() and return an np.ndarray in column-major order\n * def __array__(self): # <<<<<<<<<<<<<<\n * cdef np.npy_intp shape[1]\n * shape[0] = self._h*self._w*self._n\n */\n\n /* function exit code */\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_1);\n __Pyx_XDECREF(__pyx_t_2);\n __Pyx_XDECREF(__pyx_t_3);\n __Pyx_XDECREF(__pyx_t_4);\n __Pyx_XDECREF(__pyx_t_5);\n __Pyx_AddTraceback(\"pycocotools._mask.Masks.__array__\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n __pyx_L0:;\n __Pyx_XDECREF(__pyx_v_ndarray);\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"(tree fragment)\":1\n * def __reduce_cython__(self): # <<<<<<<<<<<<<<\n * raise TypeError(\"no default __reduce__ due to non-trivial __cinit__\")\n * def __setstate_cython__(self, __pyx_state):\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_5Masks_5__reduce_cython__(PyObject *__pyx_v_self, CYTHON_UNUSED PyObject *unused); /*proto*/\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_5Masks_5__reduce_cython__(PyObject *__pyx_v_self, CYTHON_UNUSED PyObject *unused) {\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"__reduce_cython__ (wrapper)\", 0);\n __pyx_r = __pyx_pf_11pycocotools_5_mask_5Masks_4__reduce_cython__(((struct __pyx_obj_11pycocotools_5_mask_Masks *)__pyx_v_self));\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_5Masks_4__reduce_cython__(CYTHON_UNUSED struct __pyx_obj_11pycocotools_5_mask_Masks *__pyx_v_self) {\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n PyObject *__pyx_t_1 = NULL;\n __Pyx_RefNannySetupContext(\"__reduce_cython__\", 0);\n\n /* \"(tree fragment)\":2\n * def __reduce_cython__(self):\n * raise TypeError(\"no default __reduce__ due to non-trivial __cinit__\") # <<<<<<<<<<<<<<\n * def __setstate_cython__(self, __pyx_state):\n * raise TypeError(\"no default __reduce__ due to non-trivial __cinit__\")\n */\n __pyx_t_1 = __Pyx_PyObject_Call(__pyx_builtin_TypeError, __pyx_tuple__3, NULL); if (unlikely(!__pyx_t_1)) __PYX_ERR(1, 2, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_Raise(__pyx_t_1, 0, 0, 0);\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __PYX_ERR(1, 2, __pyx_L1_error)\n\n /* \"(tree fragment)\":1\n * def __reduce_cython__(self): # <<<<<<<<<<<<<<\n * raise TypeError(\"no default __reduce__ due to non-trivial __cinit__\")\n * def __setstate_cython__(self, __pyx_state):\n */\n\n /* function exit code */\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_1);\n __Pyx_AddTraceback(\"pycocotools._mask.Masks.__reduce_cython__\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"(tree fragment)\":3\n * def __reduce_cython__(self):\n * raise TypeError(\"no default __reduce__ due to non-trivial __cinit__\")\n * def __setstate_cython__(self, __pyx_state): # <<<<<<<<<<<<<<\n * raise TypeError(\"no default __reduce__ due to non-trivial __cinit__\")\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_5Masks_7__setstate_cython__(PyObject *__pyx_v_self, PyObject *__pyx_v___pyx_state); /*proto*/\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_5Masks_7__setstate_cython__(PyObject *__pyx_v_self, PyObject *__pyx_v___pyx_state) {\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"__setstate_cython__ (wrapper)\", 0);\n __pyx_r = __pyx_pf_11pycocotools_5_mask_5Masks_6__setstate_cython__(((struct __pyx_obj_11pycocotools_5_mask_Masks *)__pyx_v_self), ((PyObject *)__pyx_v___pyx_state));\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_5Masks_6__setstate_cython__(CYTHON_UNUSED struct __pyx_obj_11pycocotools_5_mask_Masks *__pyx_v_self, CYTHON_UNUSED PyObject *__pyx_v___pyx_state) {\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n PyObject *__pyx_t_1 = NULL;\n __Pyx_RefNannySetupContext(\"__setstate_cython__\", 0);\n\n /* \"(tree fragment)\":4\n * raise TypeError(\"no default __reduce__ due to non-trivial __cinit__\")\n * def __setstate_cython__(self, __pyx_state):\n * raise TypeError(\"no default __reduce__ due to non-trivial __cinit__\") # <<<<<<<<<<<<<<\n */\n __pyx_t_1 = __Pyx_PyObject_Call(__pyx_builtin_TypeError, __pyx_tuple__4, NULL); if (unlikely(!__pyx_t_1)) __PYX_ERR(1, 4, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_Raise(__pyx_t_1, 0, 0, 0);\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __PYX_ERR(1, 4, __pyx_L1_error)\n\n /* \"(tree fragment)\":3\n * def __reduce_cython__(self):\n * raise TypeError(\"no default __reduce__ due to non-trivial __cinit__\")\n * def __setstate_cython__(self, __pyx_state): # <<<<<<<<<<<<<<\n * raise TypeError(\"no default __reduce__ due to non-trivial __cinit__\")\n */\n\n /* function exit code */\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_1);\n __Pyx_AddTraceback(\"pycocotools._mask.Masks.__setstate_cython__\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"pycocotools/_mask.pyx\":100\n * \n * # internal conversion from Python RLEs object to compressed RLE format\n * def _toString(RLEs Rs): # <<<<<<<<<<<<<<\n * cdef siz n = Rs.n\n * cdef bytes py_string\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_1_toString(PyObject *__pyx_self, PyObject *__pyx_v_Rs); /*proto*/\nstatic PyMethodDef __pyx_mdef_11pycocotools_5_mask_1_toString = {\"_toString\", (PyCFunction)__pyx_pw_11pycocotools_5_mask_1_toString, METH_O, 0};\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_1_toString(PyObject *__pyx_self, PyObject *__pyx_v_Rs) {\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"_toString (wrapper)\", 0);\n if (unlikely(!__Pyx_ArgTypeTest(((PyObject *)__pyx_v_Rs), __pyx_ptype_11pycocotools_5_mask_RLEs, 1, \"Rs\", 0))) __PYX_ERR(0, 100, __pyx_L1_error)\n __pyx_r = __pyx_pf_11pycocotools_5_mask__toString(__pyx_self, ((struct __pyx_obj_11pycocotools_5_mask_RLEs *)__pyx_v_Rs));\n\n /* function exit code */\n goto __pyx_L0;\n __pyx_L1_error:;\n __pyx_r = NULL;\n __pyx_L0:;\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask__toString(CYTHON_UNUSED PyObject *__pyx_self, struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_Rs) {\n siz __pyx_v_n;\n PyObject *__pyx_v_py_string = 0;\n char *__pyx_v_c_string;\n PyObject *__pyx_v_objs = NULL;\n siz __pyx_v_i;\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n PyObject *__pyx_t_1 = NULL;\n siz __pyx_t_2;\n siz __pyx_t_3;\n siz __pyx_t_4;\n PyObject *__pyx_t_5 = NULL;\n PyObject *__pyx_t_6 = NULL;\n PyObject *__pyx_t_7 = NULL;\n int __pyx_t_8;\n __Pyx_RefNannySetupContext(\"_toString\", 0);\n\n /* \"pycocotools/_mask.pyx\":101\n * # internal conversion from Python RLEs object to compressed RLE format\n * def _toString(RLEs Rs):\n * cdef siz n = Rs.n # <<<<<<<<<<<<<<\n * cdef bytes py_string\n * cdef char* c_string\n */\n __pyx_t_1 = __Pyx_PyObject_GetAttrStr(((PyObject *)__pyx_v_Rs), __pyx_n_s_n); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 101, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_2 = __Pyx_PyInt_As_siz(__pyx_t_1); if (unlikely((__pyx_t_2 == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 101, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __pyx_v_n = __pyx_t_2;\n\n /* \"pycocotools/_mask.pyx\":104\n * cdef bytes py_string\n * cdef char* c_string\n * objs = [] # <<<<<<<<<<<<<<\n * for i in range(n):\n * c_string = rleToString( &Rs._R[i] )\n */\n __pyx_t_1 = PyList_New(0); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 104, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_v_objs = ((PyObject*)__pyx_t_1);\n __pyx_t_1 = 0;\n\n /* \"pycocotools/_mask.pyx\":105\n * cdef char* c_string\n * objs = []\n * for i in range(n): # <<<<<<<<<<<<<<\n * c_string = rleToString( &Rs._R[i] )\n * py_string = c_string\n */\n __pyx_t_2 = __pyx_v_n;\n __pyx_t_3 = __pyx_t_2;\n for (__pyx_t_4 = 0; __pyx_t_4 < __pyx_t_3; __pyx_t_4+=1) {\n __pyx_v_i = __pyx_t_4;\n\n /* \"pycocotools/_mask.pyx\":106\n * objs = []\n * for i in range(n):\n * c_string = rleToString( &Rs._R[i] ) # <<<<<<<<<<<<<<\n * py_string = c_string\n * objs.append({\n */\n __pyx_v_c_string = rleToString(((RLE *)(&(__pyx_v_Rs->_R[__pyx_v_i]))));\n\n /* \"pycocotools/_mask.pyx\":107\n * for i in range(n):\n * c_string = rleToString( &Rs._R[i] )\n * py_string = c_string # <<<<<<<<<<<<<<\n * objs.append({\n * 'size': [Rs._R[i].h, Rs._R[i].w],\n */\n __pyx_t_1 = __Pyx_PyBytes_FromString(__pyx_v_c_string); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 107, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_XDECREF_SET(__pyx_v_py_string, ((PyObject*)__pyx_t_1));\n __pyx_t_1 = 0;\n\n /* \"pycocotools/_mask.pyx\":109\n * py_string = c_string\n * objs.append({\n * 'size': [Rs._R[i].h, Rs._R[i].w], # <<<<<<<<<<<<<<\n * 'counts': py_string\n * })\n */\n __pyx_t_1 = __Pyx_PyDict_NewPresized(2); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 109, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_5 = __Pyx_PyInt_From_siz((__pyx_v_Rs->_R[__pyx_v_i]).h); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 109, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __pyx_t_6 = __Pyx_PyInt_From_siz((__pyx_v_Rs->_R[__pyx_v_i]).w); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 109, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_6);\n __pyx_t_7 = PyList_New(2); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 109, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_7);\n __Pyx_GIVEREF(__pyx_t_5);\n PyList_SET_ITEM(__pyx_t_7, 0, __pyx_t_5);\n __Pyx_GIVEREF(__pyx_t_6);\n PyList_SET_ITEM(__pyx_t_7, 1, __pyx_t_6);\n __pyx_t_5 = 0;\n __pyx_t_6 = 0;\n if (PyDict_SetItem(__pyx_t_1, __pyx_n_s_size, __pyx_t_7) < 0) __PYX_ERR(0, 109, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;\n\n /* \"pycocotools/_mask.pyx\":110\n * objs.append({\n * 'size': [Rs._R[i].h, Rs._R[i].w],\n * 'counts': py_string # <<<<<<<<<<<<<<\n * })\n * free(c_string)\n */\n if (PyDict_SetItem(__pyx_t_1, __pyx_n_s_counts, __pyx_v_py_string) < 0) __PYX_ERR(0, 109, __pyx_L1_error)\n\n /* \"pycocotools/_mask.pyx\":108\n * c_string = rleToString( &Rs._R[i] )\n * py_string = c_string\n * objs.append({ # <<<<<<<<<<<<<<\n * 'size': [Rs._R[i].h, Rs._R[i].w],\n * 'counts': py_string\n */\n __pyx_t_8 = __Pyx_PyList_Append(__pyx_v_objs, __pyx_t_1); if (unlikely(__pyx_t_8 == ((int)-1))) __PYX_ERR(0, 108, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n\n /* \"pycocotools/_mask.pyx\":112\n * 'counts': py_string\n * })\n * free(c_string) # <<<<<<<<<<<<<<\n * return objs\n * \n */\n free(__pyx_v_c_string);\n }\n\n /* \"pycocotools/_mask.pyx\":113\n * })\n * free(c_string)\n * return objs # <<<<<<<<<<<<<<\n * \n * # internal conversion from compressed RLE format to Python RLEs object\n */\n __Pyx_XDECREF(__pyx_r);\n __Pyx_INCREF(__pyx_v_objs);\n __pyx_r = __pyx_v_objs;\n goto __pyx_L0;\n\n /* \"pycocotools/_mask.pyx\":100\n * \n * # internal conversion from Python RLEs object to compressed RLE format\n * def _toString(RLEs Rs): # <<<<<<<<<<<<<<\n * cdef siz n = Rs.n\n * cdef bytes py_string\n */\n\n /* function exit code */\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_1);\n __Pyx_XDECREF(__pyx_t_5);\n __Pyx_XDECREF(__pyx_t_6);\n __Pyx_XDECREF(__pyx_t_7);\n __Pyx_AddTraceback(\"pycocotools._mask._toString\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n __pyx_L0:;\n __Pyx_XDECREF(__pyx_v_py_string);\n __Pyx_XDECREF(__pyx_v_objs);\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"pycocotools/_mask.pyx\":116\n * \n * # internal conversion from compressed RLE format to Python RLEs object\n * def _frString(rleObjs): # <<<<<<<<<<<<<<\n * cdef siz n = len(rleObjs)\n * Rs = RLEs(n)\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_3_frString(PyObject *__pyx_self, PyObject *__pyx_v_rleObjs); /*proto*/\nstatic PyMethodDef __pyx_mdef_11pycocotools_5_mask_3_frString = {\"_frString\", (PyCFunction)__pyx_pw_11pycocotools_5_mask_3_frString, METH_O, 0};\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_3_frString(PyObject *__pyx_self, PyObject *__pyx_v_rleObjs) {\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"_frString (wrapper)\", 0);\n __pyx_r = __pyx_pf_11pycocotools_5_mask_2_frString(__pyx_self, ((PyObject *)__pyx_v_rleObjs));\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_2_frString(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_rleObjs) {\n siz __pyx_v_n;\n struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_Rs = NULL;\n PyObject *__pyx_v_py_string = 0;\n char *__pyx_v_c_string;\n PyObject *__pyx_v_i = NULL;\n PyObject *__pyx_v_obj = NULL;\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n Py_ssize_t __pyx_t_1;\n PyObject *__pyx_t_2 = NULL;\n PyObject *__pyx_t_3 = NULL;\n PyObject *(*__pyx_t_4)(PyObject *);\n PyObject *__pyx_t_5 = NULL;\n PyObject *__pyx_t_6 = NULL;\n char *__pyx_t_7;\n Py_ssize_t __pyx_t_8;\n siz __pyx_t_9;\n siz __pyx_t_10;\n __Pyx_RefNannySetupContext(\"_frString\", 0);\n\n /* \"pycocotools/_mask.pyx\":117\n * # internal conversion from compressed RLE format to Python RLEs object\n * def _frString(rleObjs):\n * cdef siz n = len(rleObjs) # <<<<<<<<<<<<<<\n * Rs = RLEs(n)\n * cdef bytes py_string\n */\n __pyx_t_1 = PyObject_Length(__pyx_v_rleObjs); if (unlikely(__pyx_t_1 == ((Py_ssize_t)-1))) __PYX_ERR(0, 117, __pyx_L1_error)\n __pyx_v_n = __pyx_t_1;\n\n /* \"pycocotools/_mask.pyx\":118\n * def _frString(rleObjs):\n * cdef siz n = len(rleObjs)\n * Rs = RLEs(n) # <<<<<<<<<<<<<<\n * cdef bytes py_string\n * cdef char* c_string\n */\n __pyx_t_2 = __Pyx_PyInt_From_siz(__pyx_v_n); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 118, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __pyx_t_3 = __Pyx_PyObject_CallOneArg(((PyObject *)__pyx_ptype_11pycocotools_5_mask_RLEs), __pyx_t_2); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 118, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n __pyx_v_Rs = ((struct __pyx_obj_11pycocotools_5_mask_RLEs *)__pyx_t_3);\n __pyx_t_3 = 0;\n\n /* \"pycocotools/_mask.pyx\":121\n * cdef bytes py_string\n * cdef char* c_string\n * for i, obj in enumerate(rleObjs): # <<<<<<<<<<<<<<\n * py_string = str(obj['counts'])\n * c_string = py_string\n */\n __Pyx_INCREF(__pyx_int_0);\n __pyx_t_3 = __pyx_int_0;\n if (likely(PyList_CheckExact(__pyx_v_rleObjs)) || PyTuple_CheckExact(__pyx_v_rleObjs)) {\n __pyx_t_2 = __pyx_v_rleObjs; __Pyx_INCREF(__pyx_t_2); __pyx_t_1 = 0;\n __pyx_t_4 = NULL;\n } else {\n __pyx_t_1 = -1; __pyx_t_2 = PyObject_GetIter(__pyx_v_rleObjs); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 121, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __pyx_t_4 = Py_TYPE(__pyx_t_2)->tp_iternext; if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 121, __pyx_L1_error)\n }\n for (;;) {\n if (likely(!__pyx_t_4)) {\n if (likely(PyList_CheckExact(__pyx_t_2))) {\n if (__pyx_t_1 >= PyList_GET_SIZE(__pyx_t_2)) break;\n #if CYTHON_ASSUME_SAFE_MACROS && !CYTHON_AVOID_BORROWED_REFS\n __pyx_t_5 = PyList_GET_ITEM(__pyx_t_2, __pyx_t_1); __Pyx_INCREF(__pyx_t_5); __pyx_t_1++; if (unlikely(0 < 0)) __PYX_ERR(0, 121, __pyx_L1_error)\n #else\n __pyx_t_5 = PySequence_ITEM(__pyx_t_2, __pyx_t_1); __pyx_t_1++; if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 121, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n #endif\n } else {\n if (__pyx_t_1 >= PyTuple_GET_SIZE(__pyx_t_2)) break;\n #if CYTHON_ASSUME_SAFE_MACROS && !CYTHON_AVOID_BORROWED_REFS\n __pyx_t_5 = PyTuple_GET_ITEM(__pyx_t_2, __pyx_t_1); __Pyx_INCREF(__pyx_t_5); __pyx_t_1++; if (unlikely(0 < 0)) __PYX_ERR(0, 121, __pyx_L1_error)\n #else\n __pyx_t_5 = PySequence_ITEM(__pyx_t_2, __pyx_t_1); __pyx_t_1++; if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 121, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n #endif\n }\n } else {\n __pyx_t_5 = __pyx_t_4(__pyx_t_2);\n if (unlikely(!__pyx_t_5)) {\n PyObject* exc_type = PyErr_Occurred();\n if (exc_type) {\n if (likely(__Pyx_PyErr_GivenExceptionMatches(exc_type, PyExc_StopIteration))) PyErr_Clear();\n else __PYX_ERR(0, 121, __pyx_L1_error)\n }\n break;\n }\n __Pyx_GOTREF(__pyx_t_5);\n }\n __Pyx_XDECREF_SET(__pyx_v_obj, __pyx_t_5);\n __pyx_t_5 = 0;\n __Pyx_INCREF(__pyx_t_3);\n __Pyx_XDECREF_SET(__pyx_v_i, __pyx_t_3);\n __pyx_t_5 = __Pyx_PyInt_AddObjC(__pyx_t_3, __pyx_int_1, 1, 0, 0); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 121, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_DECREF(__pyx_t_3);\n __pyx_t_3 = __pyx_t_5;\n __pyx_t_5 = 0;\n\n /* \"pycocotools/_mask.pyx\":122\n * cdef char* c_string\n * for i, obj in enumerate(rleObjs):\n * py_string = str(obj['counts']) # <<<<<<<<<<<<<<\n * c_string = py_string\n * rleFrString( &Rs._R[i], c_string, obj['size'][0], obj['size'][1] )\n */\n __pyx_t_5 = __Pyx_PyObject_Dict_GetItem(__pyx_v_obj, __pyx_n_s_counts); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 122, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __pyx_t_6 = __Pyx_PyObject_CallOneArg(((PyObject *)(&PyString_Type)), __pyx_t_5); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 122, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_6);\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n if (!(likely(PyBytes_CheckExact(__pyx_t_6))||((__pyx_t_6) == Py_None)||(PyErr_Format(PyExc_TypeError, \"Expected %.16s, got %.200s\", \"bytes\", Py_TYPE(__pyx_t_6)->tp_name), 0))) __PYX_ERR(0, 122, __pyx_L1_error)\n __Pyx_XDECREF_SET(__pyx_v_py_string, ((PyObject*)__pyx_t_6));\n __pyx_t_6 = 0;\n\n /* \"pycocotools/_mask.pyx\":123\n * for i, obj in enumerate(rleObjs):\n * py_string = str(obj['counts'])\n * c_string = py_string # <<<<<<<<<<<<<<\n * rleFrString( &Rs._R[i], c_string, obj['size'][0], obj['size'][1] )\n * return Rs\n */\n if (unlikely(__pyx_v_py_string == Py_None)) {\n PyErr_SetString(PyExc_TypeError, \"expected bytes, NoneType found\");\n __PYX_ERR(0, 123, __pyx_L1_error)\n }\n __pyx_t_7 = __Pyx_PyBytes_AsWritableString(__pyx_v_py_string); if (unlikely((!__pyx_t_7) && PyErr_Occurred())) __PYX_ERR(0, 123, __pyx_L1_error)\n __pyx_v_c_string = __pyx_t_7;\n\n /* \"pycocotools/_mask.pyx\":124\n * py_string = str(obj['counts'])\n * c_string = py_string\n * rleFrString( &Rs._R[i], c_string, obj['size'][0], obj['size'][1] ) # <<<<<<<<<<<<<<\n * return Rs\n * \n */\n __pyx_t_8 = __Pyx_PyIndex_AsSsize_t(__pyx_v_i); if (unlikely((__pyx_t_8 == (Py_ssize_t)-1) && PyErr_Occurred())) __PYX_ERR(0, 124, __pyx_L1_error)\n __pyx_t_6 = __Pyx_PyObject_Dict_GetItem(__pyx_v_obj, __pyx_n_s_size); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 124, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_6);\n __pyx_t_5 = __Pyx_GetItemInt(__pyx_t_6, 0, long, 1, __Pyx_PyInt_From_long, 0, 0, 1); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 124, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;\n __pyx_t_9 = __Pyx_PyInt_As_siz(__pyx_t_5); if (unlikely((__pyx_t_9 == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 124, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __pyx_t_5 = __Pyx_PyObject_Dict_GetItem(__pyx_v_obj, __pyx_n_s_size); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 124, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __pyx_t_6 = __Pyx_GetItemInt(__pyx_t_5, 1, long, 1, __Pyx_PyInt_From_long, 0, 0, 1); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 124, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_6);\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __pyx_t_10 = __Pyx_PyInt_As_siz(__pyx_t_6); if (unlikely((__pyx_t_10 == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 124, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;\n rleFrString(((RLE *)(&(__pyx_v_Rs->_R[__pyx_t_8]))), ((char *)__pyx_v_c_string), __pyx_t_9, __pyx_t_10);\n\n /* \"pycocotools/_mask.pyx\":121\n * cdef bytes py_string\n * cdef char* c_string\n * for i, obj in enumerate(rleObjs): # <<<<<<<<<<<<<<\n * py_string = str(obj['counts'])\n * c_string = py_string\n */\n }\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n\n /* \"pycocotools/_mask.pyx\":125\n * c_string = py_string\n * rleFrString( &Rs._R[i], c_string, obj['size'][0], obj['size'][1] )\n * return Rs # <<<<<<<<<<<<<<\n * \n * # encode mask to RLEs objects\n */\n __Pyx_XDECREF(__pyx_r);\n __Pyx_INCREF(((PyObject *)__pyx_v_Rs));\n __pyx_r = ((PyObject *)__pyx_v_Rs);\n goto __pyx_L0;\n\n /* \"pycocotools/_mask.pyx\":116\n * \n * # internal conversion from compressed RLE format to Python RLEs object\n * def _frString(rleObjs): # <<<<<<<<<<<<<<\n * cdef siz n = len(rleObjs)\n * Rs = RLEs(n)\n */\n\n /* function exit code */\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_2);\n __Pyx_XDECREF(__pyx_t_3);\n __Pyx_XDECREF(__pyx_t_5);\n __Pyx_XDECREF(__pyx_t_6);\n __Pyx_AddTraceback(\"pycocotools._mask._frString\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n __pyx_L0:;\n __Pyx_XDECREF((PyObject *)__pyx_v_Rs);\n __Pyx_XDECREF(__pyx_v_py_string);\n __Pyx_XDECREF(__pyx_v_i);\n __Pyx_XDECREF(__pyx_v_obj);\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"pycocotools/_mask.pyx\":129\n * # encode mask to RLEs objects\n * # list of RLE string can be generated by RLEs member function\n * def encode(np.ndarray[np.uint8_t, ndim=3, mode='fortran'] mask): # <<<<<<<<<<<<<<\n * h, w, n = mask.shape[0], mask.shape[1], mask.shape[2]\n * cdef RLEs Rs = RLEs(n)\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_5encode(PyObject *__pyx_self, PyObject *__pyx_v_mask); /*proto*/\nstatic PyMethodDef __pyx_mdef_11pycocotools_5_mask_5encode = {\"encode\", (PyCFunction)__pyx_pw_11pycocotools_5_mask_5encode, METH_O, 0};\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_5encode(PyObject *__pyx_self, PyObject *__pyx_v_mask) {\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"encode (wrapper)\", 0);\n if (unlikely(!__Pyx_ArgTypeTest(((PyObject *)__pyx_v_mask), __pyx_ptype_5numpy_ndarray, 1, \"mask\", 0))) __PYX_ERR(0, 129, __pyx_L1_error)\n __pyx_r = __pyx_pf_11pycocotools_5_mask_4encode(__pyx_self, ((PyArrayObject *)__pyx_v_mask));\n\n /* function exit code */\n goto __pyx_L0;\n __pyx_L1_error:;\n __pyx_r = NULL;\n __pyx_L0:;\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_4encode(CYTHON_UNUSED PyObject *__pyx_self, PyArrayObject *__pyx_v_mask) {\n npy_intp __pyx_v_h;\n npy_intp __pyx_v_w;\n npy_intp __pyx_v_n;\n struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_Rs = 0;\n PyObject *__pyx_v_objs = NULL;\n __Pyx_LocalBuf_ND __pyx_pybuffernd_mask;\n __Pyx_Buffer __pyx_pybuffer_mask;\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n npy_intp __pyx_t_1;\n npy_intp __pyx_t_2;\n npy_intp __pyx_t_3;\n PyObject *__pyx_t_4 = NULL;\n PyObject *__pyx_t_5 = NULL;\n PyObject *__pyx_t_6 = NULL;\n __Pyx_RefNannySetupContext(\"encode\", 0);\n __pyx_pybuffer_mask.pybuffer.buf = NULL;\n __pyx_pybuffer_mask.refcount = 0;\n __pyx_pybuffernd_mask.data = NULL;\n __pyx_pybuffernd_mask.rcbuffer = &__pyx_pybuffer_mask;\n {\n __Pyx_BufFmt_StackElem __pyx_stack[1];\n if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_mask.rcbuffer->pybuffer, (PyObject*)__pyx_v_mask, &__Pyx_TypeInfo_nn___pyx_t_5numpy_uint8_t, PyBUF_FORMAT| PyBUF_F_CONTIGUOUS, 3, 0, __pyx_stack) == -1)) __PYX_ERR(0, 129, __pyx_L1_error)\n }\n __pyx_pybuffernd_mask.diminfo[0].strides = __pyx_pybuffernd_mask.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_mask.diminfo[0].shape = __pyx_pybuffernd_mask.rcbuffer->pybuffer.shape[0]; __pyx_pybuffernd_mask.diminfo[1].strides = __pyx_pybuffernd_mask.rcbuffer->pybuffer.strides[1]; __pyx_pybuffernd_mask.diminfo[1].shape = __pyx_pybuffernd_mask.rcbuffer->pybuffer.shape[1]; __pyx_pybuffernd_mask.diminfo[2].strides = __pyx_pybuffernd_mask.rcbuffer->pybuffer.strides[2]; __pyx_pybuffernd_mask.diminfo[2].shape = __pyx_pybuffernd_mask.rcbuffer->pybuffer.shape[2];\n\n /* \"pycocotools/_mask.pyx\":130\n * # list of RLE string can be generated by RLEs member function\n * def encode(np.ndarray[np.uint8_t, ndim=3, mode='fortran'] mask):\n * h, w, n = mask.shape[0], mask.shape[1], mask.shape[2] # <<<<<<<<<<<<<<\n * cdef RLEs Rs = RLEs(n)\n * rleEncode(Rs._R,mask.data,h,w,n)\n */\n __pyx_t_1 = (__pyx_v_mask->dimensions[0]);\n __pyx_t_2 = (__pyx_v_mask->dimensions[1]);\n __pyx_t_3 = (__pyx_v_mask->dimensions[2]);\n __pyx_v_h = __pyx_t_1;\n __pyx_v_w = __pyx_t_2;\n __pyx_v_n = __pyx_t_3;\n\n /* \"pycocotools/_mask.pyx\":131\n * def encode(np.ndarray[np.uint8_t, ndim=3, mode='fortran'] mask):\n * h, w, n = mask.shape[0], mask.shape[1], mask.shape[2]\n * cdef RLEs Rs = RLEs(n) # <<<<<<<<<<<<<<\n * rleEncode(Rs._R,mask.data,h,w,n)\n * objs = _toString(Rs)\n */\n __pyx_t_4 = __Pyx_PyInt_From_Py_intptr_t(__pyx_v_n); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 131, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __pyx_t_5 = __Pyx_PyObject_CallOneArg(((PyObject *)__pyx_ptype_11pycocotools_5_mask_RLEs), __pyx_t_4); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 131, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;\n __pyx_v_Rs = ((struct __pyx_obj_11pycocotools_5_mask_RLEs *)__pyx_t_5);\n __pyx_t_5 = 0;\n\n /* \"pycocotools/_mask.pyx\":132\n * h, w, n = mask.shape[0], mask.shape[1], mask.shape[2]\n * cdef RLEs Rs = RLEs(n)\n * rleEncode(Rs._R,mask.data,h,w,n) # <<<<<<<<<<<<<<\n * objs = _toString(Rs)\n * return objs\n */\n rleEncode(__pyx_v_Rs->_R, ((byte *)__pyx_v_mask->data), __pyx_v_h, __pyx_v_w, __pyx_v_n);\n\n /* \"pycocotools/_mask.pyx\":133\n * cdef RLEs Rs = RLEs(n)\n * rleEncode(Rs._R,mask.data,h,w,n)\n * objs = _toString(Rs) # <<<<<<<<<<<<<<\n * return objs\n * \n */\n __Pyx_GetModuleGlobalName(__pyx_t_4, __pyx_n_s_toString); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 133, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __pyx_t_6 = NULL;\n if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_4))) {\n __pyx_t_6 = PyMethod_GET_SELF(__pyx_t_4);\n if (likely(__pyx_t_6)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_4);\n __Pyx_INCREF(__pyx_t_6);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_4, function);\n }\n }\n __pyx_t_5 = (__pyx_t_6) ? __Pyx_PyObject_Call2Args(__pyx_t_4, __pyx_t_6, ((PyObject *)__pyx_v_Rs)) : __Pyx_PyObject_CallOneArg(__pyx_t_4, ((PyObject *)__pyx_v_Rs));\n __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;\n if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 133, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;\n __pyx_v_objs = __pyx_t_5;\n __pyx_t_5 = 0;\n\n /* \"pycocotools/_mask.pyx\":134\n * rleEncode(Rs._R,mask.data,h,w,n)\n * objs = _toString(Rs)\n * return objs # <<<<<<<<<<<<<<\n * \n * # decode mask from compressed list of RLE string or RLEs object\n */\n __Pyx_XDECREF(__pyx_r);\n __Pyx_INCREF(__pyx_v_objs);\n __pyx_r = __pyx_v_objs;\n goto __pyx_L0;\n\n /* \"pycocotools/_mask.pyx\":129\n * # encode mask to RLEs objects\n * # list of RLE string can be generated by RLEs member function\n * def encode(np.ndarray[np.uint8_t, ndim=3, mode='fortran'] mask): # <<<<<<<<<<<<<<\n * h, w, n = mask.shape[0], mask.shape[1], mask.shape[2]\n * cdef RLEs Rs = RLEs(n)\n */\n\n /* function exit code */\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_4);\n __Pyx_XDECREF(__pyx_t_5);\n __Pyx_XDECREF(__pyx_t_6);\n { PyObject *__pyx_type, *__pyx_value, *__pyx_tb;\n __Pyx_PyThreadState_declare\n __Pyx_PyThreadState_assign\n __Pyx_ErrFetch(&__pyx_type, &__pyx_value, &__pyx_tb);\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_mask.rcbuffer->pybuffer);\n __Pyx_ErrRestore(__pyx_type, __pyx_value, __pyx_tb);}\n __Pyx_AddTraceback(\"pycocotools._mask.encode\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n goto __pyx_L2;\n __pyx_L0:;\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_mask.rcbuffer->pybuffer);\n __pyx_L2:;\n __Pyx_XDECREF((PyObject *)__pyx_v_Rs);\n __Pyx_XDECREF(__pyx_v_objs);\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"pycocotools/_mask.pyx\":137\n * \n * # decode mask from compressed list of RLE string or RLEs object\n * def decode(rleObjs): # <<<<<<<<<<<<<<\n * cdef RLEs Rs = _frString(rleObjs)\n * h, w, n = Rs._R[0].h, Rs._R[0].w, Rs._n\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_7decode(PyObject *__pyx_self, PyObject *__pyx_v_rleObjs); /*proto*/\nstatic PyMethodDef __pyx_mdef_11pycocotools_5_mask_7decode = {\"decode\", (PyCFunction)__pyx_pw_11pycocotools_5_mask_7decode, METH_O, 0};\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_7decode(PyObject *__pyx_self, PyObject *__pyx_v_rleObjs) {\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"decode (wrapper)\", 0);\n __pyx_r = __pyx_pf_11pycocotools_5_mask_6decode(__pyx_self, ((PyObject *)__pyx_v_rleObjs));\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_6decode(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_rleObjs) {\n struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_Rs = 0;\n siz __pyx_v_h;\n siz __pyx_v_w;\n siz __pyx_v_n;\n struct __pyx_obj_11pycocotools_5_mask_Masks *__pyx_v_masks = NULL;\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n PyObject *__pyx_t_1 = NULL;\n PyObject *__pyx_t_2 = NULL;\n PyObject *__pyx_t_3 = NULL;\n siz __pyx_t_4;\n siz __pyx_t_5;\n siz __pyx_t_6;\n PyObject *__pyx_t_7 = NULL;\n __Pyx_RefNannySetupContext(\"decode\", 0);\n\n /* \"pycocotools/_mask.pyx\":138\n * # decode mask from compressed list of RLE string or RLEs object\n * def decode(rleObjs):\n * cdef RLEs Rs = _frString(rleObjs) # <<<<<<<<<<<<<<\n * h, w, n = Rs._R[0].h, Rs._R[0].w, Rs._n\n * masks = Masks(h, w, n)\n */\n __Pyx_GetModuleGlobalName(__pyx_t_2, __pyx_n_s_frString); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 138, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __pyx_t_3 = NULL;\n if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_2))) {\n __pyx_t_3 = PyMethod_GET_SELF(__pyx_t_2);\n if (likely(__pyx_t_3)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_2);\n __Pyx_INCREF(__pyx_t_3);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_2, function);\n }\n }\n __pyx_t_1 = (__pyx_t_3) ? __Pyx_PyObject_Call2Args(__pyx_t_2, __pyx_t_3, __pyx_v_rleObjs) : __Pyx_PyObject_CallOneArg(__pyx_t_2, __pyx_v_rleObjs);\n __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;\n if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 138, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n if (!(likely(((__pyx_t_1) == Py_None) || likely(__Pyx_TypeTest(__pyx_t_1, __pyx_ptype_11pycocotools_5_mask_RLEs))))) __PYX_ERR(0, 138, __pyx_L1_error)\n __pyx_v_Rs = ((struct __pyx_obj_11pycocotools_5_mask_RLEs *)__pyx_t_1);\n __pyx_t_1 = 0;\n\n /* \"pycocotools/_mask.pyx\":139\n * def decode(rleObjs):\n * cdef RLEs Rs = _frString(rleObjs)\n * h, w, n = Rs._R[0].h, Rs._R[0].w, Rs._n # <<<<<<<<<<<<<<\n * masks = Masks(h, w, n)\n * rleDecode( Rs._R, masks._mask, n );\n */\n __pyx_t_4 = (__pyx_v_Rs->_R[0]).h;\n __pyx_t_5 = (__pyx_v_Rs->_R[0]).w;\n __pyx_t_6 = __pyx_v_Rs->_n;\n __pyx_v_h = __pyx_t_4;\n __pyx_v_w = __pyx_t_5;\n __pyx_v_n = __pyx_t_6;\n\n /* \"pycocotools/_mask.pyx\":140\n * cdef RLEs Rs = _frString(rleObjs)\n * h, w, n = Rs._R[0].h, Rs._R[0].w, Rs._n\n * masks = Masks(h, w, n) # <<<<<<<<<<<<<<\n * rleDecode( Rs._R, masks._mask, n );\n * return np.array(masks)\n */\n __pyx_t_1 = __Pyx_PyInt_From_siz(__pyx_v_h); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 140, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_2 = __Pyx_PyInt_From_siz(__pyx_v_w); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 140, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __pyx_t_3 = __Pyx_PyInt_From_siz(__pyx_v_n); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 140, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __pyx_t_7 = PyTuple_New(3); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 140, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_7);\n __Pyx_GIVEREF(__pyx_t_1);\n PyTuple_SET_ITEM(__pyx_t_7, 0, __pyx_t_1);\n __Pyx_GIVEREF(__pyx_t_2);\n PyTuple_SET_ITEM(__pyx_t_7, 1, __pyx_t_2);\n __Pyx_GIVEREF(__pyx_t_3);\n PyTuple_SET_ITEM(__pyx_t_7, 2, __pyx_t_3);\n __pyx_t_1 = 0;\n __pyx_t_2 = 0;\n __pyx_t_3 = 0;\n __pyx_t_3 = __Pyx_PyObject_Call(((PyObject *)__pyx_ptype_11pycocotools_5_mask_Masks), __pyx_t_7, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 140, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;\n __pyx_v_masks = ((struct __pyx_obj_11pycocotools_5_mask_Masks *)__pyx_t_3);\n __pyx_t_3 = 0;\n\n /* \"pycocotools/_mask.pyx\":141\n * h, w, n = Rs._R[0].h, Rs._R[0].w, Rs._n\n * masks = Masks(h, w, n)\n * rleDecode( Rs._R, masks._mask, n ); # <<<<<<<<<<<<<<\n * return np.array(masks)\n * \n */\n rleDecode(((RLE *)__pyx_v_Rs->_R), __pyx_v_masks->_mask, __pyx_v_n);\n\n /* \"pycocotools/_mask.pyx\":142\n * masks = Masks(h, w, n)\n * rleDecode( Rs._R, masks._mask, n );\n * return np.array(masks) # <<<<<<<<<<<<<<\n * \n * def merge(rleObjs, bint intersect=0):\n */\n __Pyx_XDECREF(__pyx_r);\n __Pyx_GetModuleGlobalName(__pyx_t_7, __pyx_n_s_np); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 142, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_7);\n __pyx_t_2 = __Pyx_PyObject_GetAttrStr(__pyx_t_7, __pyx_n_s_array); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 142, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;\n __pyx_t_7 = NULL;\n if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_2))) {\n __pyx_t_7 = PyMethod_GET_SELF(__pyx_t_2);\n if (likely(__pyx_t_7)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_2);\n __Pyx_INCREF(__pyx_t_7);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_2, function);\n }\n }\n __pyx_t_3 = (__pyx_t_7) ? __Pyx_PyObject_Call2Args(__pyx_t_2, __pyx_t_7, ((PyObject *)__pyx_v_masks)) : __Pyx_PyObject_CallOneArg(__pyx_t_2, ((PyObject *)__pyx_v_masks));\n __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;\n if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 142, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n __pyx_r = __pyx_t_3;\n __pyx_t_3 = 0;\n goto __pyx_L0;\n\n /* \"pycocotools/_mask.pyx\":137\n * \n * # decode mask from compressed list of RLE string or RLEs object\n * def decode(rleObjs): # <<<<<<<<<<<<<<\n * cdef RLEs Rs = _frString(rleObjs)\n * h, w, n = Rs._R[0].h, Rs._R[0].w, Rs._n\n */\n\n /* function exit code */\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_1);\n __Pyx_XDECREF(__pyx_t_2);\n __Pyx_XDECREF(__pyx_t_3);\n __Pyx_XDECREF(__pyx_t_7);\n __Pyx_AddTraceback(\"pycocotools._mask.decode\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n __pyx_L0:;\n __Pyx_XDECREF((PyObject *)__pyx_v_Rs);\n __Pyx_XDECREF((PyObject *)__pyx_v_masks);\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"pycocotools/_mask.pyx\":144\n * return np.array(masks)\n * \n * def merge(rleObjs, bint intersect=0): # <<<<<<<<<<<<<<\n * cdef RLEs Rs = _frString(rleObjs)\n * cdef RLEs R = RLEs(1)\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_9merge(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds); /*proto*/\nstatic PyMethodDef __pyx_mdef_11pycocotools_5_mask_9merge = {\"merge\", (PyCFunction)(void*)(PyCFunctionWithKeywords)__pyx_pw_11pycocotools_5_mask_9merge, METH_VARARGS|METH_KEYWORDS, 0};\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_9merge(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds) {\n PyObject *__pyx_v_rleObjs = 0;\n int __pyx_v_intersect;\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"merge (wrapper)\", 0);\n {\n static PyObject **__pyx_pyargnames[] = {&__pyx_n_s_rleObjs,&__pyx_n_s_intersect,0};\n PyObject* values[2] = {0,0};\n if (unlikely(__pyx_kwds)) {\n Py_ssize_t kw_args;\n const Py_ssize_t pos_args = PyTuple_GET_SIZE(__pyx_args);\n switch (pos_args) {\n case 2: values[1] = PyTuple_GET_ITEM(__pyx_args, 1);\n CYTHON_FALLTHROUGH;\n case 1: values[0] = PyTuple_GET_ITEM(__pyx_args, 0);\n CYTHON_FALLTHROUGH;\n case 0: break;\n default: goto __pyx_L5_argtuple_error;\n }\n kw_args = PyDict_Size(__pyx_kwds);\n switch (pos_args) {\n case 0:\n if (likely((values[0] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_rleObjs)) != 0)) kw_args--;\n else goto __pyx_L5_argtuple_error;\n CYTHON_FALLTHROUGH;\n case 1:\n if (kw_args > 0) {\n PyObject* value = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_intersect);\n if (value) { values[1] = value; kw_args--; }\n }\n }\n if (unlikely(kw_args > 0)) {\n if (unlikely(__Pyx_ParseOptionalKeywords(__pyx_kwds, __pyx_pyargnames, 0, values, pos_args, \"merge\") < 0)) __PYX_ERR(0, 144, __pyx_L3_error)\n }\n } else {\n switch (PyTuple_GET_SIZE(__pyx_args)) {\n case 2: values[1] = PyTuple_GET_ITEM(__pyx_args, 1);\n CYTHON_FALLTHROUGH;\n case 1: values[0] = PyTuple_GET_ITEM(__pyx_args, 0);\n break;\n default: goto __pyx_L5_argtuple_error;\n }\n }\n __pyx_v_rleObjs = values[0];\n if (values[1]) {\n __pyx_v_intersect = __Pyx_PyObject_IsTrue(values[1]); if (unlikely((__pyx_v_intersect == (int)-1) && PyErr_Occurred())) __PYX_ERR(0, 144, __pyx_L3_error)\n } else {\n __pyx_v_intersect = ((int)0);\n }\n }\n goto __pyx_L4_argument_unpacking_done;\n __pyx_L5_argtuple_error:;\n __Pyx_RaiseArgtupleInvalid(\"merge\", 0, 1, 2, PyTuple_GET_SIZE(__pyx_args)); __PYX_ERR(0, 144, __pyx_L3_error)\n __pyx_L3_error:;\n __Pyx_AddTraceback(\"pycocotools._mask.merge\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __Pyx_RefNannyFinishContext();\n return NULL;\n __pyx_L4_argument_unpacking_done:;\n __pyx_r = __pyx_pf_11pycocotools_5_mask_8merge(__pyx_self, __pyx_v_rleObjs, __pyx_v_intersect);\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_8merge(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_rleObjs, int __pyx_v_intersect) {\n struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_Rs = 0;\n struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_R = 0;\n PyObject *__pyx_v_obj = NULL;\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n PyObject *__pyx_t_1 = NULL;\n PyObject *__pyx_t_2 = NULL;\n PyObject *__pyx_t_3 = NULL;\n __Pyx_RefNannySetupContext(\"merge\", 0);\n\n /* \"pycocotools/_mask.pyx\":145\n * \n * def merge(rleObjs, bint intersect=0):\n * cdef RLEs Rs = _frString(rleObjs) # <<<<<<<<<<<<<<\n * cdef RLEs R = RLEs(1)\n * rleMerge(Rs._R, R._R, Rs._n, intersect)\n */\n __Pyx_GetModuleGlobalName(__pyx_t_2, __pyx_n_s_frString); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 145, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __pyx_t_3 = NULL;\n if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_2))) {\n __pyx_t_3 = PyMethod_GET_SELF(__pyx_t_2);\n if (likely(__pyx_t_3)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_2);\n __Pyx_INCREF(__pyx_t_3);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_2, function);\n }\n }\n __pyx_t_1 = (__pyx_t_3) ? __Pyx_PyObject_Call2Args(__pyx_t_2, __pyx_t_3, __pyx_v_rleObjs) : __Pyx_PyObject_CallOneArg(__pyx_t_2, __pyx_v_rleObjs);\n __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;\n if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 145, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n if (!(likely(((__pyx_t_1) == Py_None) || likely(__Pyx_TypeTest(__pyx_t_1, __pyx_ptype_11pycocotools_5_mask_RLEs))))) __PYX_ERR(0, 145, __pyx_L1_error)\n __pyx_v_Rs = ((struct __pyx_obj_11pycocotools_5_mask_RLEs *)__pyx_t_1);\n __pyx_t_1 = 0;\n\n /* \"pycocotools/_mask.pyx\":146\n * def merge(rleObjs, bint intersect=0):\n * cdef RLEs Rs = _frString(rleObjs)\n * cdef RLEs R = RLEs(1) # <<<<<<<<<<<<<<\n * rleMerge(Rs._R, R._R, Rs._n, intersect)\n * obj = _toString(R)[0]\n */\n __pyx_t_1 = __Pyx_PyObject_Call(((PyObject *)__pyx_ptype_11pycocotools_5_mask_RLEs), __pyx_tuple__5, NULL); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 146, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_v_R = ((struct __pyx_obj_11pycocotools_5_mask_RLEs *)__pyx_t_1);\n __pyx_t_1 = 0;\n\n /* \"pycocotools/_mask.pyx\":147\n * cdef RLEs Rs = _frString(rleObjs)\n * cdef RLEs R = RLEs(1)\n * rleMerge(Rs._R, R._R, Rs._n, intersect) # <<<<<<<<<<<<<<\n * obj = _toString(R)[0]\n * return obj\n */\n rleMerge(((RLE *)__pyx_v_Rs->_R), ((RLE *)__pyx_v_R->_R), ((siz)__pyx_v_Rs->_n), __pyx_v_intersect);\n\n /* \"pycocotools/_mask.pyx\":148\n * cdef RLEs R = RLEs(1)\n * rleMerge(Rs._R, R._R, Rs._n, intersect)\n * obj = _toString(R)[0] # <<<<<<<<<<<<<<\n * return obj\n * \n */\n __Pyx_GetModuleGlobalName(__pyx_t_2, __pyx_n_s_toString); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 148, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __pyx_t_3 = NULL;\n if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_2))) {\n __pyx_t_3 = PyMethod_GET_SELF(__pyx_t_2);\n if (likely(__pyx_t_3)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_2);\n __Pyx_INCREF(__pyx_t_3);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_2, function);\n }\n }\n __pyx_t_1 = (__pyx_t_3) ? __Pyx_PyObject_Call2Args(__pyx_t_2, __pyx_t_3, ((PyObject *)__pyx_v_R)) : __Pyx_PyObject_CallOneArg(__pyx_t_2, ((PyObject *)__pyx_v_R));\n __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;\n if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 148, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n __pyx_t_2 = __Pyx_GetItemInt(__pyx_t_1, 0, long, 1, __Pyx_PyInt_From_long, 0, 0, 1); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 148, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __pyx_v_obj = __pyx_t_2;\n __pyx_t_2 = 0;\n\n /* \"pycocotools/_mask.pyx\":149\n * rleMerge(Rs._R, R._R, Rs._n, intersect)\n * obj = _toString(R)[0]\n * return obj # <<<<<<<<<<<<<<\n * \n * def area(rleObjs):\n */\n __Pyx_XDECREF(__pyx_r);\n __Pyx_INCREF(__pyx_v_obj);\n __pyx_r = __pyx_v_obj;\n goto __pyx_L0;\n\n /* \"pycocotools/_mask.pyx\":144\n * return np.array(masks)\n * \n * def merge(rleObjs, bint intersect=0): # <<<<<<<<<<<<<<\n * cdef RLEs Rs = _frString(rleObjs)\n * cdef RLEs R = RLEs(1)\n */\n\n /* function exit code */\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_1);\n __Pyx_XDECREF(__pyx_t_2);\n __Pyx_XDECREF(__pyx_t_3);\n __Pyx_AddTraceback(\"pycocotools._mask.merge\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n __pyx_L0:;\n __Pyx_XDECREF((PyObject *)__pyx_v_Rs);\n __Pyx_XDECREF((PyObject *)__pyx_v_R);\n __Pyx_XDECREF(__pyx_v_obj);\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"pycocotools/_mask.pyx\":151\n * return obj\n * \n * def area(rleObjs): # <<<<<<<<<<<<<<\n * cdef RLEs Rs = _frString(rleObjs)\n * cdef uint* _a = malloc(Rs._n* sizeof(uint))\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_11area(PyObject *__pyx_self, PyObject *__pyx_v_rleObjs); /*proto*/\nstatic PyMethodDef __pyx_mdef_11pycocotools_5_mask_11area = {\"area\", (PyCFunction)__pyx_pw_11pycocotools_5_mask_11area, METH_O, 0};\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_11area(PyObject *__pyx_self, PyObject *__pyx_v_rleObjs) {\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"area (wrapper)\", 0);\n __pyx_r = __pyx_pf_11pycocotools_5_mask_10area(__pyx_self, ((PyObject *)__pyx_v_rleObjs));\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_10area(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_rleObjs) {\n struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_Rs = 0;\n uint *__pyx_v__a;\n npy_intp __pyx_v_shape[1];\n PyObject *__pyx_v_a = NULL;\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n PyObject *__pyx_t_1 = NULL;\n PyObject *__pyx_t_2 = NULL;\n PyObject *__pyx_t_3 = NULL;\n PyObject *__pyx_t_4 = NULL;\n PyObject *__pyx_t_5 = NULL;\n __Pyx_RefNannySetupContext(\"area\", 0);\n\n /* \"pycocotools/_mask.pyx\":152\n * \n * def area(rleObjs):\n * cdef RLEs Rs = _frString(rleObjs) # <<<<<<<<<<<<<<\n * cdef uint* _a = malloc(Rs._n* sizeof(uint))\n * rleArea(Rs._R, Rs._n, _a)\n */\n __Pyx_GetModuleGlobalName(__pyx_t_2, __pyx_n_s_frString); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 152, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __pyx_t_3 = NULL;\n if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_2))) {\n __pyx_t_3 = PyMethod_GET_SELF(__pyx_t_2);\n if (likely(__pyx_t_3)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_2);\n __Pyx_INCREF(__pyx_t_3);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_2, function);\n }\n }\n __pyx_t_1 = (__pyx_t_3) ? __Pyx_PyObject_Call2Args(__pyx_t_2, __pyx_t_3, __pyx_v_rleObjs) : __Pyx_PyObject_CallOneArg(__pyx_t_2, __pyx_v_rleObjs);\n __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;\n if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 152, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n if (!(likely(((__pyx_t_1) == Py_None) || likely(__Pyx_TypeTest(__pyx_t_1, __pyx_ptype_11pycocotools_5_mask_RLEs))))) __PYX_ERR(0, 152, __pyx_L1_error)\n __pyx_v_Rs = ((struct __pyx_obj_11pycocotools_5_mask_RLEs *)__pyx_t_1);\n __pyx_t_1 = 0;\n\n /* \"pycocotools/_mask.pyx\":153\n * def area(rleObjs):\n * cdef RLEs Rs = _frString(rleObjs)\n * cdef uint* _a = malloc(Rs._n* sizeof(uint)) # <<<<<<<<<<<<<<\n * rleArea(Rs._R, Rs._n, _a)\n * cdef np.npy_intp shape[1]\n */\n __pyx_v__a = ((uint *)malloc((__pyx_v_Rs->_n * (sizeof(unsigned int)))));\n\n /* \"pycocotools/_mask.pyx\":154\n * cdef RLEs Rs = _frString(rleObjs)\n * cdef uint* _a = malloc(Rs._n* sizeof(uint))\n * rleArea(Rs._R, Rs._n, _a) # <<<<<<<<<<<<<<\n * cdef np.npy_intp shape[1]\n * shape[0] = Rs._n\n */\n rleArea(__pyx_v_Rs->_R, __pyx_v_Rs->_n, __pyx_v__a);\n\n /* \"pycocotools/_mask.pyx\":156\n * rleArea(Rs._R, Rs._n, _a)\n * cdef np.npy_intp shape[1]\n * shape[0] = Rs._n # <<<<<<<<<<<<<<\n * a = np.array((Rs._n, ), dtype=np.uint8)\n * a = np.PyArray_SimpleNewFromData(1, shape, np.NPY_UINT32, _a)\n */\n (__pyx_v_shape[0]) = ((npy_intp)__pyx_v_Rs->_n);\n\n /* \"pycocotools/_mask.pyx\":157\n * cdef np.npy_intp shape[1]\n * shape[0] = Rs._n\n * a = np.array((Rs._n, ), dtype=np.uint8) # <<<<<<<<<<<<<<\n * a = np.PyArray_SimpleNewFromData(1, shape, np.NPY_UINT32, _a)\n * PyArray_ENABLEFLAGS(a, np.NPY_OWNDATA)\n */\n __Pyx_GetModuleGlobalName(__pyx_t_1, __pyx_n_s_np); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 157, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_2 = __Pyx_PyObject_GetAttrStr(__pyx_t_1, __pyx_n_s_array); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 157, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __pyx_t_1 = __Pyx_PyInt_From_siz(__pyx_v_Rs->_n); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 157, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_3 = PyTuple_New(1); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 157, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_GIVEREF(__pyx_t_1);\n PyTuple_SET_ITEM(__pyx_t_3, 0, __pyx_t_1);\n __pyx_t_1 = 0;\n __pyx_t_1 = PyTuple_New(1); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 157, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_GIVEREF(__pyx_t_3);\n PyTuple_SET_ITEM(__pyx_t_1, 0, __pyx_t_3);\n __pyx_t_3 = 0;\n __pyx_t_3 = __Pyx_PyDict_NewPresized(1); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 157, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_GetModuleGlobalName(__pyx_t_4, __pyx_n_s_np); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 157, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __pyx_t_5 = __Pyx_PyObject_GetAttrStr(__pyx_t_4, __pyx_n_s_uint8); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 157, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;\n if (PyDict_SetItem(__pyx_t_3, __pyx_n_s_dtype, __pyx_t_5) < 0) __PYX_ERR(0, 157, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __pyx_t_5 = __Pyx_PyObject_Call(__pyx_t_2, __pyx_t_1, __pyx_t_3); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 157, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __pyx_v_a = __pyx_t_5;\n __pyx_t_5 = 0;\n\n /* \"pycocotools/_mask.pyx\":158\n * shape[0] = Rs._n\n * a = np.array((Rs._n, ), dtype=np.uint8)\n * a = np.PyArray_SimpleNewFromData(1, shape, np.NPY_UINT32, _a) # <<<<<<<<<<<<<<\n * PyArray_ENABLEFLAGS(a, np.NPY_OWNDATA)\n * return a\n */\n __pyx_t_5 = PyArray_SimpleNewFromData(1, __pyx_v_shape, NPY_UINT32, __pyx_v__a); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 158, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_DECREF_SET(__pyx_v_a, __pyx_t_5);\n __pyx_t_5 = 0;\n\n /* \"pycocotools/_mask.pyx\":159\n * a = np.array((Rs._n, ), dtype=np.uint8)\n * a = np.PyArray_SimpleNewFromData(1, shape, np.NPY_UINT32, _a)\n * PyArray_ENABLEFLAGS(a, np.NPY_OWNDATA) # <<<<<<<<<<<<<<\n * return a\n * \n */\n if (!(likely(((__pyx_v_a) == Py_None) || likely(__Pyx_TypeTest(__pyx_v_a, __pyx_ptype_5numpy_ndarray))))) __PYX_ERR(0, 159, __pyx_L1_error)\n PyArray_ENABLEFLAGS(((PyArrayObject *)__pyx_v_a), NPY_OWNDATA);\n\n /* \"pycocotools/_mask.pyx\":160\n * a = np.PyArray_SimpleNewFromData(1, shape, np.NPY_UINT32, _a)\n * PyArray_ENABLEFLAGS(a, np.NPY_OWNDATA)\n * return a # <<<<<<<<<<<<<<\n * \n * # iou computation. support function overload (RLEs-RLEs and bbox-bbox).\n */\n __Pyx_XDECREF(__pyx_r);\n __Pyx_INCREF(__pyx_v_a);\n __pyx_r = __pyx_v_a;\n goto __pyx_L0;\n\n /* \"pycocotools/_mask.pyx\":151\n * return obj\n * \n * def area(rleObjs): # <<<<<<<<<<<<<<\n * cdef RLEs Rs = _frString(rleObjs)\n * cdef uint* _a = malloc(Rs._n* sizeof(uint))\n */\n\n /* function exit code */\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_1);\n __Pyx_XDECREF(__pyx_t_2);\n __Pyx_XDECREF(__pyx_t_3);\n __Pyx_XDECREF(__pyx_t_4);\n __Pyx_XDECREF(__pyx_t_5);\n __Pyx_AddTraceback(\"pycocotools._mask.area\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n __pyx_L0:;\n __Pyx_XDECREF((PyObject *)__pyx_v_Rs);\n __Pyx_XDECREF(__pyx_v_a);\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"pycocotools/_mask.pyx\":163\n * \n * # iou computation. support function overload (RLEs-RLEs and bbox-bbox).\n * def iou( dt, gt, pyiscrowd ): # <<<<<<<<<<<<<<\n * def _preproc(objs):\n * if len(objs) == 0:\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_13iou(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds); /*proto*/\nstatic PyMethodDef __pyx_mdef_11pycocotools_5_mask_13iou = {\"iou\", (PyCFunction)(void*)(PyCFunctionWithKeywords)__pyx_pw_11pycocotools_5_mask_13iou, METH_VARARGS|METH_KEYWORDS, 0};\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_13iou(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds) {\n PyObject *__pyx_v_dt = 0;\n PyObject *__pyx_v_gt = 0;\n PyObject *__pyx_v_pyiscrowd = 0;\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"iou (wrapper)\", 0);\n {\n static PyObject **__pyx_pyargnames[] = {&__pyx_n_s_dt,&__pyx_n_s_gt,&__pyx_n_s_pyiscrowd,0};\n PyObject* values[3] = {0,0,0};\n if (unlikely(__pyx_kwds)) {\n Py_ssize_t kw_args;\n const Py_ssize_t pos_args = PyTuple_GET_SIZE(__pyx_args);\n switch (pos_args) {\n case 3: values[2] = PyTuple_GET_ITEM(__pyx_args, 2);\n CYTHON_FALLTHROUGH;\n case 2: values[1] = PyTuple_GET_ITEM(__pyx_args, 1);\n CYTHON_FALLTHROUGH;\n case 1: values[0] = PyTuple_GET_ITEM(__pyx_args, 0);\n CYTHON_FALLTHROUGH;\n case 0: break;\n default: goto __pyx_L5_argtuple_error;\n }\n kw_args = PyDict_Size(__pyx_kwds);\n switch (pos_args) {\n case 0:\n if (likely((values[0] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_dt)) != 0)) kw_args--;\n else goto __pyx_L5_argtuple_error;\n CYTHON_FALLTHROUGH;\n case 1:\n if (likely((values[1] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_gt)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"iou\", 1, 3, 3, 1); __PYX_ERR(0, 163, __pyx_L3_error)\n }\n CYTHON_FALLTHROUGH;\n case 2:\n if (likely((values[2] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_pyiscrowd)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"iou\", 1, 3, 3, 2); __PYX_ERR(0, 163, __pyx_L3_error)\n }\n }\n if (unlikely(kw_args > 0)) {\n if (unlikely(__Pyx_ParseOptionalKeywords(__pyx_kwds, __pyx_pyargnames, 0, values, pos_args, \"iou\") < 0)) __PYX_ERR(0, 163, __pyx_L3_error)\n }\n } else if (PyTuple_GET_SIZE(__pyx_args) != 3) {\n goto __pyx_L5_argtuple_error;\n } else {\n values[0] = PyTuple_GET_ITEM(__pyx_args, 0);\n values[1] = PyTuple_GET_ITEM(__pyx_args, 1);\n values[2] = PyTuple_GET_ITEM(__pyx_args, 2);\n }\n __pyx_v_dt = values[0];\n __pyx_v_gt = values[1];\n __pyx_v_pyiscrowd = values[2];\n }\n goto __pyx_L4_argument_unpacking_done;\n __pyx_L5_argtuple_error:;\n __Pyx_RaiseArgtupleInvalid(\"iou\", 1, 3, 3, PyTuple_GET_SIZE(__pyx_args)); __PYX_ERR(0, 163, __pyx_L3_error)\n __pyx_L3_error:;\n __Pyx_AddTraceback(\"pycocotools._mask.iou\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __Pyx_RefNannyFinishContext();\n return NULL;\n __pyx_L4_argument_unpacking_done:;\n __pyx_r = __pyx_pf_11pycocotools_5_mask_12iou(__pyx_self, __pyx_v_dt, __pyx_v_gt, __pyx_v_pyiscrowd);\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"pycocotools/_mask.pyx\":164\n * # iou computation. support function overload (RLEs-RLEs and bbox-bbox).\n * def iou( dt, gt, pyiscrowd ):\n * def _preproc(objs): # <<<<<<<<<<<<<<\n * if len(objs) == 0:\n * return objs\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_3iou_1_preproc(PyObject *__pyx_self, PyObject *__pyx_v_objs); /*proto*/\nstatic PyMethodDef __pyx_mdef_11pycocotools_5_mask_3iou_1_preproc = {\"_preproc\", (PyCFunction)__pyx_pw_11pycocotools_5_mask_3iou_1_preproc, METH_O, 0};\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_3iou_1_preproc(PyObject *__pyx_self, PyObject *__pyx_v_objs) {\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"_preproc (wrapper)\", 0);\n __pyx_r = __pyx_pf_11pycocotools_5_mask_3iou__preproc(__pyx_self, ((PyObject *)__pyx_v_objs));\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_3iou__preproc(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_objs) {\n PyObject *__pyx_v_isbox = NULL;\n PyObject *__pyx_v_isrle = NULL;\n PyObject *__pyx_v_obj = NULL;\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n Py_ssize_t __pyx_t_1;\n int __pyx_t_2;\n PyObject *__pyx_t_3 = NULL;\n PyObject *__pyx_t_4 = NULL;\n PyObject *__pyx_t_5 = NULL;\n PyObject *__pyx_t_6 = NULL;\n int __pyx_t_7;\n int __pyx_t_8;\n PyObject *__pyx_t_9 = NULL;\n PyObject *__pyx_t_10 = NULL;\n PyObject *(*__pyx_t_11)(PyObject *);\n PyObject *__pyx_t_12 = NULL;\n Py_ssize_t __pyx_t_13;\n PyObject *__pyx_t_14 = NULL;\n __Pyx_RefNannySetupContext(\"_preproc\", 0);\n __Pyx_INCREF(__pyx_v_objs);\n\n /* \"pycocotools/_mask.pyx\":165\n * def iou( dt, gt, pyiscrowd ):\n * def _preproc(objs):\n * if len(objs) == 0: # <<<<<<<<<<<<<<\n * return objs\n * if type(objs) == np.ndarray:\n */\n __pyx_t_1 = PyObject_Length(__pyx_v_objs); if (unlikely(__pyx_t_1 == ((Py_ssize_t)-1))) __PYX_ERR(0, 165, __pyx_L1_error)\n __pyx_t_2 = ((__pyx_t_1 == 0) != 0);\n if (__pyx_t_2) {\n\n /* \"pycocotools/_mask.pyx\":166\n * def _preproc(objs):\n * if len(objs) == 0:\n * return objs # <<<<<<<<<<<<<<\n * if type(objs) == np.ndarray:\n * if len(objs.shape) == 1:\n */\n __Pyx_XDECREF(__pyx_r);\n __Pyx_INCREF(__pyx_v_objs);\n __pyx_r = __pyx_v_objs;\n goto __pyx_L0;\n\n /* \"pycocotools/_mask.pyx\":165\n * def iou( dt, gt, pyiscrowd ):\n * def _preproc(objs):\n * if len(objs) == 0: # <<<<<<<<<<<<<<\n * return objs\n * if type(objs) == np.ndarray:\n */\n }\n\n /* \"pycocotools/_mask.pyx\":167\n * if len(objs) == 0:\n * return objs\n * if type(objs) == np.ndarray: # <<<<<<<<<<<<<<\n * if len(objs.shape) == 1:\n * objs = objs.reshape((objs[0], 1))\n */\n __pyx_t_3 = PyObject_RichCompare(((PyObject *)Py_TYPE(__pyx_v_objs)), ((PyObject *)__pyx_ptype_5numpy_ndarray), Py_EQ); __Pyx_XGOTREF(__pyx_t_3); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 167, __pyx_L1_error)\n __pyx_t_2 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely(__pyx_t_2 < 0)) __PYX_ERR(0, 167, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n if (__pyx_t_2) {\n\n /* \"pycocotools/_mask.pyx\":168\n * return objs\n * if type(objs) == np.ndarray:\n * if len(objs.shape) == 1: # <<<<<<<<<<<<<<\n * objs = objs.reshape((objs[0], 1))\n * # check if it's Nx4 bbox\n */\n __pyx_t_3 = __Pyx_PyObject_GetAttrStr(__pyx_v_objs, __pyx_n_s_shape); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 168, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __pyx_t_1 = PyObject_Length(__pyx_t_3); if (unlikely(__pyx_t_1 == ((Py_ssize_t)-1))) __PYX_ERR(0, 168, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __pyx_t_2 = ((__pyx_t_1 == 1) != 0);\n if (__pyx_t_2) {\n\n /* \"pycocotools/_mask.pyx\":169\n * if type(objs) == np.ndarray:\n * if len(objs.shape) == 1:\n * objs = objs.reshape((objs[0], 1)) # <<<<<<<<<<<<<<\n * # check if it's Nx4 bbox\n * if not len(objs.shape) == 2 or not objs.shape[1] == 4:\n */\n __pyx_t_4 = __Pyx_PyObject_GetAttrStr(__pyx_v_objs, __pyx_n_s_reshape); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 169, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __pyx_t_5 = __Pyx_GetItemInt(__pyx_v_objs, 0, long, 1, __Pyx_PyInt_From_long, 0, 0, 1); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 169, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __pyx_t_6 = PyTuple_New(2); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 169, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_6);\n __Pyx_GIVEREF(__pyx_t_5);\n PyTuple_SET_ITEM(__pyx_t_6, 0, __pyx_t_5);\n __Pyx_INCREF(__pyx_int_1);\n __Pyx_GIVEREF(__pyx_int_1);\n PyTuple_SET_ITEM(__pyx_t_6, 1, __pyx_int_1);\n __pyx_t_5 = 0;\n __pyx_t_5 = NULL;\n if (CYTHON_UNPACK_METHODS && likely(PyMethod_Check(__pyx_t_4))) {\n __pyx_t_5 = PyMethod_GET_SELF(__pyx_t_4);\n if (likely(__pyx_t_5)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_4);\n __Pyx_INCREF(__pyx_t_5);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_4, function);\n }\n }\n __pyx_t_3 = (__pyx_t_5) ? __Pyx_PyObject_Call2Args(__pyx_t_4, __pyx_t_5, __pyx_t_6) : __Pyx_PyObject_CallOneArg(__pyx_t_4, __pyx_t_6);\n __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;\n __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;\n if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 169, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;\n __Pyx_DECREF_SET(__pyx_v_objs, __pyx_t_3);\n __pyx_t_3 = 0;\n\n /* \"pycocotools/_mask.pyx\":168\n * return objs\n * if type(objs) == np.ndarray:\n * if len(objs.shape) == 1: # <<<<<<<<<<<<<<\n * objs = objs.reshape((objs[0], 1))\n * # check if it's Nx4 bbox\n */\n }\n\n /* \"pycocotools/_mask.pyx\":171\n * objs = objs.reshape((objs[0], 1))\n * # check if it's Nx4 bbox\n * if not len(objs.shape) == 2 or not objs.shape[1] == 4: # <<<<<<<<<<<<<<\n * raise Exception('numpy ndarray input is only for *bounding boxes* and should have Nx4 dimension')\n * objs = objs.astype(np.double)\n */\n __pyx_t_3 = __Pyx_PyObject_GetAttrStr(__pyx_v_objs, __pyx_n_s_shape); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 171, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __pyx_t_1 = PyObject_Length(__pyx_t_3); if (unlikely(__pyx_t_1 == ((Py_ssize_t)-1))) __PYX_ERR(0, 171, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __pyx_t_7 = ((!((__pyx_t_1 == 2) != 0)) != 0);\n if (!__pyx_t_7) {\n } else {\n __pyx_t_2 = __pyx_t_7;\n goto __pyx_L7_bool_binop_done;\n }\n __pyx_t_3 = __Pyx_PyObject_GetAttrStr(__pyx_v_objs, __pyx_n_s_shape); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 171, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __pyx_t_4 = __Pyx_GetItemInt(__pyx_t_3, 1, long, 1, __Pyx_PyInt_From_long, 0, 0, 1); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 171, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __pyx_t_3 = __Pyx_PyInt_EqObjC(__pyx_t_4, __pyx_int_4, 4, 0); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 171, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;\n __pyx_t_7 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely(__pyx_t_7 < 0)) __PYX_ERR(0, 171, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __pyx_t_8 = ((!__pyx_t_7) != 0);\n __pyx_t_2 = __pyx_t_8;\n __pyx_L7_bool_binop_done:;\n if (unlikely(__pyx_t_2)) {\n\n /* \"pycocotools/_mask.pyx\":172\n * # check if it's Nx4 bbox\n * if not len(objs.shape) == 2 or not objs.shape[1] == 4:\n * raise Exception('numpy ndarray input is only for *bounding boxes* and should have Nx4 dimension') # <<<<<<<<<<<<<<\n * objs = objs.astype(np.double)\n * elif type(objs) == list:\n */\n __pyx_t_3 = __Pyx_PyObject_Call(((PyObject *)(&((PyTypeObject*)PyExc_Exception)[0])), __pyx_tuple__6, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 172, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_Raise(__pyx_t_3, 0, 0, 0);\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __PYX_ERR(0, 172, __pyx_L1_error)\n\n /* \"pycocotools/_mask.pyx\":171\n * objs = objs.reshape((objs[0], 1))\n * # check if it's Nx4 bbox\n * if not len(objs.shape) == 2 or not objs.shape[1] == 4: # <<<<<<<<<<<<<<\n * raise Exception('numpy ndarray input is only for *bounding boxes* and should have Nx4 dimension')\n * objs = objs.astype(np.double)\n */\n }\n\n /* \"pycocotools/_mask.pyx\":173\n * if not len(objs.shape) == 2 or not objs.shape[1] == 4:\n * raise Exception('numpy ndarray input is only for *bounding boxes* and should have Nx4 dimension')\n * objs = objs.astype(np.double) # <<<<<<<<<<<<<<\n * elif type(objs) == list:\n * # check if list is in box format and convert it to np.ndarray\n */\n __pyx_t_4 = __Pyx_PyObject_GetAttrStr(__pyx_v_objs, __pyx_n_s_astype); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 173, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __Pyx_GetModuleGlobalName(__pyx_t_6, __pyx_n_s_np); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 173, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_6);\n __pyx_t_5 = __Pyx_PyObject_GetAttrStr(__pyx_t_6, __pyx_n_s_double); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 173, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;\n __pyx_t_6 = NULL;\n if (CYTHON_UNPACK_METHODS && likely(PyMethod_Check(__pyx_t_4))) {\n __pyx_t_6 = PyMethod_GET_SELF(__pyx_t_4);\n if (likely(__pyx_t_6)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_4);\n __Pyx_INCREF(__pyx_t_6);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_4, function);\n }\n }\n __pyx_t_3 = (__pyx_t_6) ? __Pyx_PyObject_Call2Args(__pyx_t_4, __pyx_t_6, __pyx_t_5) : __Pyx_PyObject_CallOneArg(__pyx_t_4, __pyx_t_5);\n __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 173, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;\n __Pyx_DECREF_SET(__pyx_v_objs, __pyx_t_3);\n __pyx_t_3 = 0;\n\n /* \"pycocotools/_mask.pyx\":167\n * if len(objs) == 0:\n * return objs\n * if type(objs) == np.ndarray: # <<<<<<<<<<<<<<\n * if len(objs.shape) == 1:\n * objs = objs.reshape((objs[0], 1))\n */\n goto __pyx_L4;\n }\n\n /* \"pycocotools/_mask.pyx\":174\n * raise Exception('numpy ndarray input is only for *bounding boxes* and should have Nx4 dimension')\n * objs = objs.astype(np.double)\n * elif type(objs) == list: # <<<<<<<<<<<<<<\n * # check if list is in box format and convert it to np.ndarray\n * isbox = np.all(np.array([(len(obj)==4) and ((type(obj)==list) or (type(obj)==np.ndarray)) for obj in objs]))\n */\n __pyx_t_3 = PyObject_RichCompare(((PyObject *)Py_TYPE(__pyx_v_objs)), ((PyObject *)(&PyList_Type)), Py_EQ); __Pyx_XGOTREF(__pyx_t_3); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 174, __pyx_L1_error)\n __pyx_t_2 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely(__pyx_t_2 < 0)) __PYX_ERR(0, 174, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n if (likely(__pyx_t_2)) {\n\n /* \"pycocotools/_mask.pyx\":176\n * elif type(objs) == list:\n * # check if list is in box format and convert it to np.ndarray\n * isbox = np.all(np.array([(len(obj)==4) and ((type(obj)==list) or (type(obj)==np.ndarray)) for obj in objs])) # <<<<<<<<<<<<<<\n * isrle = np.all(np.array([type(obj) == dict for obj in objs]))\n * if isbox:\n */\n __Pyx_GetModuleGlobalName(__pyx_t_4, __pyx_n_s_np); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 176, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __pyx_t_5 = __Pyx_PyObject_GetAttrStr(__pyx_t_4, __pyx_n_s_all); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 176, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;\n __Pyx_GetModuleGlobalName(__pyx_t_6, __pyx_n_s_np); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 176, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_6);\n __pyx_t_9 = __Pyx_PyObject_GetAttrStr(__pyx_t_6, __pyx_n_s_array); if (unlikely(!__pyx_t_9)) __PYX_ERR(0, 176, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_9);\n __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;\n __pyx_t_6 = PyList_New(0); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 176, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_6);\n if (likely(PyList_CheckExact(__pyx_v_objs)) || PyTuple_CheckExact(__pyx_v_objs)) {\n __pyx_t_10 = __pyx_v_objs; __Pyx_INCREF(__pyx_t_10); __pyx_t_1 = 0;\n __pyx_t_11 = NULL;\n } else {\n __pyx_t_1 = -1; __pyx_t_10 = PyObject_GetIter(__pyx_v_objs); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 176, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_10);\n __pyx_t_11 = Py_TYPE(__pyx_t_10)->tp_iternext; if (unlikely(!__pyx_t_11)) __PYX_ERR(0, 176, __pyx_L1_error)\n }\n for (;;) {\n if (likely(!__pyx_t_11)) {\n if (likely(PyList_CheckExact(__pyx_t_10))) {\n if (__pyx_t_1 >= PyList_GET_SIZE(__pyx_t_10)) break;\n #if CYTHON_ASSUME_SAFE_MACROS && !CYTHON_AVOID_BORROWED_REFS\n __pyx_t_12 = PyList_GET_ITEM(__pyx_t_10, __pyx_t_1); __Pyx_INCREF(__pyx_t_12); __pyx_t_1++; if (unlikely(0 < 0)) __PYX_ERR(0, 176, __pyx_L1_error)\n #else\n __pyx_t_12 = PySequence_ITEM(__pyx_t_10, __pyx_t_1); __pyx_t_1++; if (unlikely(!__pyx_t_12)) __PYX_ERR(0, 176, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_12);\n #endif\n } else {\n if (__pyx_t_1 >= PyTuple_GET_SIZE(__pyx_t_10)) break;\n #if CYTHON_ASSUME_SAFE_MACROS && !CYTHON_AVOID_BORROWED_REFS\n __pyx_t_12 = PyTuple_GET_ITEM(__pyx_t_10, __pyx_t_1); __Pyx_INCREF(__pyx_t_12); __pyx_t_1++; if (unlikely(0 < 0)) __PYX_ERR(0, 176, __pyx_L1_error)\n #else\n __pyx_t_12 = PySequence_ITEM(__pyx_t_10, __pyx_t_1); __pyx_t_1++; if (unlikely(!__pyx_t_12)) __PYX_ERR(0, 176, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_12);\n #endif\n }\n } else {\n __pyx_t_12 = __pyx_t_11(__pyx_t_10);\n if (unlikely(!__pyx_t_12)) {\n PyObject* exc_type = PyErr_Occurred();\n if (exc_type) {\n if (likely(__Pyx_PyErr_GivenExceptionMatches(exc_type, PyExc_StopIteration))) PyErr_Clear();\n else __PYX_ERR(0, 176, __pyx_L1_error)\n }\n break;\n }\n __Pyx_GOTREF(__pyx_t_12);\n }\n __Pyx_XDECREF_SET(__pyx_v_obj, __pyx_t_12);\n __pyx_t_12 = 0;\n __pyx_t_13 = PyObject_Length(__pyx_v_obj); if (unlikely(__pyx_t_13 == ((Py_ssize_t)-1))) __PYX_ERR(0, 176, __pyx_L1_error)\n __pyx_t_2 = (__pyx_t_13 == 4);\n if (__pyx_t_2) {\n } else {\n __pyx_t_14 = __Pyx_PyBool_FromLong(__pyx_t_2); if (unlikely(!__pyx_t_14)) __PYX_ERR(0, 176, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_14);\n __pyx_t_12 = __pyx_t_14;\n __pyx_t_14 = 0;\n goto __pyx_L11_bool_binop_done;\n }\n __pyx_t_14 = PyObject_RichCompare(((PyObject *)Py_TYPE(__pyx_v_obj)), ((PyObject *)(&PyList_Type)), Py_EQ); __Pyx_XGOTREF(__pyx_t_14); if (unlikely(!__pyx_t_14)) __PYX_ERR(0, 176, __pyx_L1_error)\n __pyx_t_2 = __Pyx_PyObject_IsTrue(__pyx_t_14); if (unlikely(__pyx_t_2 < 0)) __PYX_ERR(0, 176, __pyx_L1_error)\n if (!__pyx_t_2) {\n __Pyx_DECREF(__pyx_t_14); __pyx_t_14 = 0;\n } else {\n __Pyx_INCREF(__pyx_t_14);\n __pyx_t_12 = __pyx_t_14;\n __Pyx_DECREF(__pyx_t_14); __pyx_t_14 = 0;\n goto __pyx_L11_bool_binop_done;\n }\n __pyx_t_14 = PyObject_RichCompare(((PyObject *)Py_TYPE(__pyx_v_obj)), ((PyObject *)__pyx_ptype_5numpy_ndarray), Py_EQ); __Pyx_XGOTREF(__pyx_t_14); if (unlikely(!__pyx_t_14)) __PYX_ERR(0, 176, __pyx_L1_error)\n __Pyx_INCREF(__pyx_t_14);\n __pyx_t_12 = __pyx_t_14;\n __Pyx_DECREF(__pyx_t_14); __pyx_t_14 = 0;\n __pyx_L11_bool_binop_done:;\n if (unlikely(__Pyx_ListComp_Append(__pyx_t_6, (PyObject*)__pyx_t_12))) __PYX_ERR(0, 176, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_12); __pyx_t_12 = 0;\n }\n __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;\n __pyx_t_10 = NULL;\n if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_9))) {\n __pyx_t_10 = PyMethod_GET_SELF(__pyx_t_9);\n if (likely(__pyx_t_10)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_9);\n __Pyx_INCREF(__pyx_t_10);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_9, function);\n }\n }\n __pyx_t_4 = (__pyx_t_10) ? __Pyx_PyObject_Call2Args(__pyx_t_9, __pyx_t_10, __pyx_t_6) : __Pyx_PyObject_CallOneArg(__pyx_t_9, __pyx_t_6);\n __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;\n __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;\n if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 176, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __Pyx_DECREF(__pyx_t_9); __pyx_t_9 = 0;\n __pyx_t_9 = NULL;\n if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_5))) {\n __pyx_t_9 = PyMethod_GET_SELF(__pyx_t_5);\n if (likely(__pyx_t_9)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_5);\n __Pyx_INCREF(__pyx_t_9);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_5, function);\n }\n }\n __pyx_t_3 = (__pyx_t_9) ? __Pyx_PyObject_Call2Args(__pyx_t_5, __pyx_t_9, __pyx_t_4) : __Pyx_PyObject_CallOneArg(__pyx_t_5, __pyx_t_4);\n __Pyx_XDECREF(__pyx_t_9); __pyx_t_9 = 0;\n __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;\n if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 176, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __pyx_v_isbox = __pyx_t_3;\n __pyx_t_3 = 0;\n\n /* \"pycocotools/_mask.pyx\":177\n * # check if list is in box format and convert it to np.ndarray\n * isbox = np.all(np.array([(len(obj)==4) and ((type(obj)==list) or (type(obj)==np.ndarray)) for obj in objs]))\n * isrle = np.all(np.array([type(obj) == dict for obj in objs])) # <<<<<<<<<<<<<<\n * if isbox:\n * objs = np.array(objs, dtype=np.double)\n */\n __Pyx_GetModuleGlobalName(__pyx_t_5, __pyx_n_s_np); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 177, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __pyx_t_4 = __Pyx_PyObject_GetAttrStr(__pyx_t_5, __pyx_n_s_all); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 177, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __Pyx_GetModuleGlobalName(__pyx_t_9, __pyx_n_s_np); if (unlikely(!__pyx_t_9)) __PYX_ERR(0, 177, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_9);\n __pyx_t_6 = __Pyx_PyObject_GetAttrStr(__pyx_t_9, __pyx_n_s_array); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 177, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_6);\n __Pyx_DECREF(__pyx_t_9); __pyx_t_9 = 0;\n __pyx_t_9 = PyList_New(0); if (unlikely(!__pyx_t_9)) __PYX_ERR(0, 177, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_9);\n if (likely(PyList_CheckExact(__pyx_v_objs)) || PyTuple_CheckExact(__pyx_v_objs)) {\n __pyx_t_10 = __pyx_v_objs; __Pyx_INCREF(__pyx_t_10); __pyx_t_1 = 0;\n __pyx_t_11 = NULL;\n } else {\n __pyx_t_1 = -1; __pyx_t_10 = PyObject_GetIter(__pyx_v_objs); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 177, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_10);\n __pyx_t_11 = Py_TYPE(__pyx_t_10)->tp_iternext; if (unlikely(!__pyx_t_11)) __PYX_ERR(0, 177, __pyx_L1_error)\n }\n for (;;) {\n if (likely(!__pyx_t_11)) {\n if (likely(PyList_CheckExact(__pyx_t_10))) {\n if (__pyx_t_1 >= PyList_GET_SIZE(__pyx_t_10)) break;\n #if CYTHON_ASSUME_SAFE_MACROS && !CYTHON_AVOID_BORROWED_REFS\n __pyx_t_12 = PyList_GET_ITEM(__pyx_t_10, __pyx_t_1); __Pyx_INCREF(__pyx_t_12); __pyx_t_1++; if (unlikely(0 < 0)) __PYX_ERR(0, 177, __pyx_L1_error)\n #else\n __pyx_t_12 = PySequence_ITEM(__pyx_t_10, __pyx_t_1); __pyx_t_1++; if (unlikely(!__pyx_t_12)) __PYX_ERR(0, 177, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_12);\n #endif\n } else {\n if (__pyx_t_1 >= PyTuple_GET_SIZE(__pyx_t_10)) break;\n #if CYTHON_ASSUME_SAFE_MACROS && !CYTHON_AVOID_BORROWED_REFS\n __pyx_t_12 = PyTuple_GET_ITEM(__pyx_t_10, __pyx_t_1); __Pyx_INCREF(__pyx_t_12); __pyx_t_1++; if (unlikely(0 < 0)) __PYX_ERR(0, 177, __pyx_L1_error)\n #else\n __pyx_t_12 = PySequence_ITEM(__pyx_t_10, __pyx_t_1); __pyx_t_1++; if (unlikely(!__pyx_t_12)) __PYX_ERR(0, 177, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_12);\n #endif\n }\n } else {\n __pyx_t_12 = __pyx_t_11(__pyx_t_10);\n if (unlikely(!__pyx_t_12)) {\n PyObject* exc_type = PyErr_Occurred();\n if (exc_type) {\n if (likely(__Pyx_PyErr_GivenExceptionMatches(exc_type, PyExc_StopIteration))) PyErr_Clear();\n else __PYX_ERR(0, 177, __pyx_L1_error)\n }\n break;\n }\n __Pyx_GOTREF(__pyx_t_12);\n }\n __Pyx_XDECREF_SET(__pyx_v_obj, __pyx_t_12);\n __pyx_t_12 = 0;\n __pyx_t_12 = PyObject_RichCompare(((PyObject *)Py_TYPE(__pyx_v_obj)), ((PyObject *)(&PyDict_Type)), Py_EQ); __Pyx_XGOTREF(__pyx_t_12); if (unlikely(!__pyx_t_12)) __PYX_ERR(0, 177, __pyx_L1_error)\n if (unlikely(__Pyx_ListComp_Append(__pyx_t_9, (PyObject*)__pyx_t_12))) __PYX_ERR(0, 177, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_12); __pyx_t_12 = 0;\n }\n __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;\n __pyx_t_10 = NULL;\n if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_6))) {\n __pyx_t_10 = PyMethod_GET_SELF(__pyx_t_6);\n if (likely(__pyx_t_10)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_6);\n __Pyx_INCREF(__pyx_t_10);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_6, function);\n }\n }\n __pyx_t_5 = (__pyx_t_10) ? __Pyx_PyObject_Call2Args(__pyx_t_6, __pyx_t_10, __pyx_t_9) : __Pyx_PyObject_CallOneArg(__pyx_t_6, __pyx_t_9);\n __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;\n __Pyx_DECREF(__pyx_t_9); __pyx_t_9 = 0;\n if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 177, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;\n __pyx_t_6 = NULL;\n if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_4))) {\n __pyx_t_6 = PyMethod_GET_SELF(__pyx_t_4);\n if (likely(__pyx_t_6)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_4);\n __Pyx_INCREF(__pyx_t_6);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_4, function);\n }\n }\n __pyx_t_3 = (__pyx_t_6) ? __Pyx_PyObject_Call2Args(__pyx_t_4, __pyx_t_6, __pyx_t_5) : __Pyx_PyObject_CallOneArg(__pyx_t_4, __pyx_t_5);\n __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 177, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;\n __pyx_v_isrle = __pyx_t_3;\n __pyx_t_3 = 0;\n\n /* \"pycocotools/_mask.pyx\":178\n * isbox = np.all(np.array([(len(obj)==4) and ((type(obj)==list) or (type(obj)==np.ndarray)) for obj in objs]))\n * isrle = np.all(np.array([type(obj) == dict for obj in objs]))\n * if isbox: # <<<<<<<<<<<<<<\n * objs = np.array(objs, dtype=np.double)\n * if len(objs.shape) == 1:\n */\n __pyx_t_2 = __Pyx_PyObject_IsTrue(__pyx_v_isbox); if (unlikely(__pyx_t_2 < 0)) __PYX_ERR(0, 178, __pyx_L1_error)\n if (__pyx_t_2) {\n\n /* \"pycocotools/_mask.pyx\":179\n * isrle = np.all(np.array([type(obj) == dict for obj in objs]))\n * if isbox:\n * objs = np.array(objs, dtype=np.double) # <<<<<<<<<<<<<<\n * if len(objs.shape) == 1:\n * objs = objs.reshape((1,objs.shape[0]))\n */\n __Pyx_GetModuleGlobalName(__pyx_t_3, __pyx_n_s_np); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 179, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __pyx_t_4 = __Pyx_PyObject_GetAttrStr(__pyx_t_3, __pyx_n_s_array); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 179, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __pyx_t_3 = PyTuple_New(1); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 179, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_INCREF(__pyx_v_objs);\n __Pyx_GIVEREF(__pyx_v_objs);\n PyTuple_SET_ITEM(__pyx_t_3, 0, __pyx_v_objs);\n __pyx_t_5 = __Pyx_PyDict_NewPresized(1); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 179, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_GetModuleGlobalName(__pyx_t_6, __pyx_n_s_np); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 179, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_6);\n __pyx_t_9 = __Pyx_PyObject_GetAttrStr(__pyx_t_6, __pyx_n_s_double); if (unlikely(!__pyx_t_9)) __PYX_ERR(0, 179, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_9);\n __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;\n if (PyDict_SetItem(__pyx_t_5, __pyx_n_s_dtype, __pyx_t_9) < 0) __PYX_ERR(0, 179, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_9); __pyx_t_9 = 0;\n __pyx_t_9 = __Pyx_PyObject_Call(__pyx_t_4, __pyx_t_3, __pyx_t_5); if (unlikely(!__pyx_t_9)) __PYX_ERR(0, 179, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_9);\n __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __Pyx_DECREF_SET(__pyx_v_objs, __pyx_t_9);\n __pyx_t_9 = 0;\n\n /* \"pycocotools/_mask.pyx\":180\n * if isbox:\n * objs = np.array(objs, dtype=np.double)\n * if len(objs.shape) == 1: # <<<<<<<<<<<<<<\n * objs = objs.reshape((1,objs.shape[0]))\n * elif isrle:\n */\n __pyx_t_9 = __Pyx_PyObject_GetAttrStr(__pyx_v_objs, __pyx_n_s_shape); if (unlikely(!__pyx_t_9)) __PYX_ERR(0, 180, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_9);\n __pyx_t_1 = PyObject_Length(__pyx_t_9); if (unlikely(__pyx_t_1 == ((Py_ssize_t)-1))) __PYX_ERR(0, 180, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_9); __pyx_t_9 = 0;\n __pyx_t_2 = ((__pyx_t_1 == 1) != 0);\n if (__pyx_t_2) {\n\n /* \"pycocotools/_mask.pyx\":181\n * objs = np.array(objs, dtype=np.double)\n * if len(objs.shape) == 1:\n * objs = objs.reshape((1,objs.shape[0])) # <<<<<<<<<<<<<<\n * elif isrle:\n * objs = _frString(objs)\n */\n __pyx_t_5 = __Pyx_PyObject_GetAttrStr(__pyx_v_objs, __pyx_n_s_reshape); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 181, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __pyx_t_3 = __Pyx_PyObject_GetAttrStr(__pyx_v_objs, __pyx_n_s_shape); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 181, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __pyx_t_4 = __Pyx_GetItemInt(__pyx_t_3, 0, long, 1, __Pyx_PyInt_From_long, 0, 0, 1); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 181, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __pyx_t_3 = PyTuple_New(2); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 181, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_INCREF(__pyx_int_1);\n __Pyx_GIVEREF(__pyx_int_1);\n PyTuple_SET_ITEM(__pyx_t_3, 0, __pyx_int_1);\n __Pyx_GIVEREF(__pyx_t_4);\n PyTuple_SET_ITEM(__pyx_t_3, 1, __pyx_t_4);\n __pyx_t_4 = 0;\n __pyx_t_4 = NULL;\n if (CYTHON_UNPACK_METHODS && likely(PyMethod_Check(__pyx_t_5))) {\n __pyx_t_4 = PyMethod_GET_SELF(__pyx_t_5);\n if (likely(__pyx_t_4)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_5);\n __Pyx_INCREF(__pyx_t_4);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_5, function);\n }\n }\n __pyx_t_9 = (__pyx_t_4) ? __Pyx_PyObject_Call2Args(__pyx_t_5, __pyx_t_4, __pyx_t_3) : __Pyx_PyObject_CallOneArg(__pyx_t_5, __pyx_t_3);\n __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n if (unlikely(!__pyx_t_9)) __PYX_ERR(0, 181, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_9);\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __Pyx_DECREF_SET(__pyx_v_objs, __pyx_t_9);\n __pyx_t_9 = 0;\n\n /* \"pycocotools/_mask.pyx\":180\n * if isbox:\n * objs = np.array(objs, dtype=np.double)\n * if len(objs.shape) == 1: # <<<<<<<<<<<<<<\n * objs = objs.reshape((1,objs.shape[0]))\n * elif isrle:\n */\n }\n\n /* \"pycocotools/_mask.pyx\":178\n * isbox = np.all(np.array([(len(obj)==4) and ((type(obj)==list) or (type(obj)==np.ndarray)) for obj in objs]))\n * isrle = np.all(np.array([type(obj) == dict for obj in objs]))\n * if isbox: # <<<<<<<<<<<<<<\n * objs = np.array(objs, dtype=np.double)\n * if len(objs.shape) == 1:\n */\n goto __pyx_L16;\n }\n\n /* \"pycocotools/_mask.pyx\":182\n * if len(objs.shape) == 1:\n * objs = objs.reshape((1,objs.shape[0]))\n * elif isrle: # <<<<<<<<<<<<<<\n * objs = _frString(objs)\n * else:\n */\n __pyx_t_2 = __Pyx_PyObject_IsTrue(__pyx_v_isrle); if (unlikely(__pyx_t_2 < 0)) __PYX_ERR(0, 182, __pyx_L1_error)\n if (likely(__pyx_t_2)) {\n\n /* \"pycocotools/_mask.pyx\":183\n * objs = objs.reshape((1,objs.shape[0]))\n * elif isrle:\n * objs = _frString(objs) # <<<<<<<<<<<<<<\n * else:\n * raise Exception('list input can be bounding box (Nx4) or RLEs ([RLE])')\n */\n __Pyx_GetModuleGlobalName(__pyx_t_5, __pyx_n_s_frString); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 183, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __pyx_t_3 = NULL;\n if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_5))) {\n __pyx_t_3 = PyMethod_GET_SELF(__pyx_t_5);\n if (likely(__pyx_t_3)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_5);\n __Pyx_INCREF(__pyx_t_3);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_5, function);\n }\n }\n __pyx_t_9 = (__pyx_t_3) ? __Pyx_PyObject_Call2Args(__pyx_t_5, __pyx_t_3, __pyx_v_objs) : __Pyx_PyObject_CallOneArg(__pyx_t_5, __pyx_v_objs);\n __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;\n if (unlikely(!__pyx_t_9)) __PYX_ERR(0, 183, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_9);\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __Pyx_DECREF_SET(__pyx_v_objs, __pyx_t_9);\n __pyx_t_9 = 0;\n\n /* \"pycocotools/_mask.pyx\":182\n * if len(objs.shape) == 1:\n * objs = objs.reshape((1,objs.shape[0]))\n * elif isrle: # <<<<<<<<<<<<<<\n * objs = _frString(objs)\n * else:\n */\n goto __pyx_L16;\n }\n\n /* \"pycocotools/_mask.pyx\":185\n * objs = _frString(objs)\n * else:\n * raise Exception('list input can be bounding box (Nx4) or RLEs ([RLE])') # <<<<<<<<<<<<<<\n * else:\n * raise Exception('unrecognized type. The following type: RLEs (rle), np.ndarray (box), and list (box) are supported.')\n */\n /*else*/ {\n __pyx_t_9 = __Pyx_PyObject_Call(((PyObject *)(&((PyTypeObject*)PyExc_Exception)[0])), __pyx_tuple__7, NULL); if (unlikely(!__pyx_t_9)) __PYX_ERR(0, 185, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_9);\n __Pyx_Raise(__pyx_t_9, 0, 0, 0);\n __Pyx_DECREF(__pyx_t_9); __pyx_t_9 = 0;\n __PYX_ERR(0, 185, __pyx_L1_error)\n }\n __pyx_L16:;\n\n /* \"pycocotools/_mask.pyx\":174\n * raise Exception('numpy ndarray input is only for *bounding boxes* and should have Nx4 dimension')\n * objs = objs.astype(np.double)\n * elif type(objs) == list: # <<<<<<<<<<<<<<\n * # check if list is in box format and convert it to np.ndarray\n * isbox = np.all(np.array([(len(obj)==4) and ((type(obj)==list) or (type(obj)==np.ndarray)) for obj in objs]))\n */\n goto __pyx_L4;\n }\n\n /* \"pycocotools/_mask.pyx\":187\n * raise Exception('list input can be bounding box (Nx4) or RLEs ([RLE])')\n * else:\n * raise Exception('unrecognized type. The following type: RLEs (rle), np.ndarray (box), and list (box) are supported.') # <<<<<<<<<<<<<<\n * return objs\n * def _rleIou(RLEs dt, RLEs gt, np.ndarray[np.uint8_t, ndim=1] iscrowd, siz m, siz n, np.ndarray[np.double_t, ndim=1] _iou):\n */\n /*else*/ {\n __pyx_t_9 = __Pyx_PyObject_Call(((PyObject *)(&((PyTypeObject*)PyExc_Exception)[0])), __pyx_tuple__8, NULL); if (unlikely(!__pyx_t_9)) __PYX_ERR(0, 187, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_9);\n __Pyx_Raise(__pyx_t_9, 0, 0, 0);\n __Pyx_DECREF(__pyx_t_9); __pyx_t_9 = 0;\n __PYX_ERR(0, 187, __pyx_L1_error)\n }\n __pyx_L4:;\n\n /* \"pycocotools/_mask.pyx\":188\n * else:\n * raise Exception('unrecognized type. The following type: RLEs (rle), np.ndarray (box), and list (box) are supported.')\n * return objs # <<<<<<<<<<<<<<\n * def _rleIou(RLEs dt, RLEs gt, np.ndarray[np.uint8_t, ndim=1] iscrowd, siz m, siz n, np.ndarray[np.double_t, ndim=1] _iou):\n * rleIou( dt._R, gt._R, m, n, iscrowd.data, _iou.data )\n */\n __Pyx_XDECREF(__pyx_r);\n __Pyx_INCREF(__pyx_v_objs);\n __pyx_r = __pyx_v_objs;\n goto __pyx_L0;\n\n /* \"pycocotools/_mask.pyx\":164\n * # iou computation. support function overload (RLEs-RLEs and bbox-bbox).\n * def iou( dt, gt, pyiscrowd ):\n * def _preproc(objs): # <<<<<<<<<<<<<<\n * if len(objs) == 0:\n * return objs\n */\n\n /* function exit code */\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_3);\n __Pyx_XDECREF(__pyx_t_4);\n __Pyx_XDECREF(__pyx_t_5);\n __Pyx_XDECREF(__pyx_t_6);\n __Pyx_XDECREF(__pyx_t_9);\n __Pyx_XDECREF(__pyx_t_10);\n __Pyx_XDECREF(__pyx_t_12);\n __Pyx_XDECREF(__pyx_t_14);\n __Pyx_AddTraceback(\"pycocotools._mask.iou._preproc\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n __pyx_L0:;\n __Pyx_XDECREF(__pyx_v_isbox);\n __Pyx_XDECREF(__pyx_v_isrle);\n __Pyx_XDECREF(__pyx_v_obj);\n __Pyx_XDECREF(__pyx_v_objs);\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"pycocotools/_mask.pyx\":189\n * raise Exception('unrecognized type. The following type: RLEs (rle), np.ndarray (box), and list (box) are supported.')\n * return objs\n * def _rleIou(RLEs dt, RLEs gt, np.ndarray[np.uint8_t, ndim=1] iscrowd, siz m, siz n, np.ndarray[np.double_t, ndim=1] _iou): # <<<<<<<<<<<<<<\n * rleIou( dt._R, gt._R, m, n, iscrowd.data, _iou.data )\n * def _bbIou(np.ndarray[np.double_t, ndim=2] dt, np.ndarray[np.double_t, ndim=2] gt, np.ndarray[np.uint8_t, ndim=1] iscrowd, siz m, siz n, np.ndarray[np.double_t, ndim=1] _iou):\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_3iou_3_rleIou(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds); /*proto*/\nstatic PyMethodDef __pyx_mdef_11pycocotools_5_mask_3iou_3_rleIou = {\"_rleIou\", (PyCFunction)(void*)(PyCFunctionWithKeywords)__pyx_pw_11pycocotools_5_mask_3iou_3_rleIou, METH_VARARGS|METH_KEYWORDS, 0};\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_3iou_3_rleIou(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds) {\n struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_dt = 0;\n struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_gt = 0;\n PyArrayObject *__pyx_v_iscrowd = 0;\n siz __pyx_v_m;\n siz __pyx_v_n;\n PyArrayObject *__pyx_v__iou = 0;\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"_rleIou (wrapper)\", 0);\n {\n static PyObject **__pyx_pyargnames[] = {&__pyx_n_s_dt,&__pyx_n_s_gt,&__pyx_n_s_iscrowd,&__pyx_n_s_m,&__pyx_n_s_n,&__pyx_n_s_iou,0};\n PyObject* values[6] = {0,0,0,0,0,0};\n if (unlikely(__pyx_kwds)) {\n Py_ssize_t kw_args;\n const Py_ssize_t pos_args = PyTuple_GET_SIZE(__pyx_args);\n switch (pos_args) {\n case 6: values[5] = PyTuple_GET_ITEM(__pyx_args, 5);\n CYTHON_FALLTHROUGH;\n case 5: values[4] = PyTuple_GET_ITEM(__pyx_args, 4);\n CYTHON_FALLTHROUGH;\n case 4: values[3] = PyTuple_GET_ITEM(__pyx_args, 3);\n CYTHON_FALLTHROUGH;\n case 3: values[2] = PyTuple_GET_ITEM(__pyx_args, 2);\n CYTHON_FALLTHROUGH;\n case 2: values[1] = PyTuple_GET_ITEM(__pyx_args, 1);\n CYTHON_FALLTHROUGH;\n case 1: values[0] = PyTuple_GET_ITEM(__pyx_args, 0);\n CYTHON_FALLTHROUGH;\n case 0: break;\n default: goto __pyx_L5_argtuple_error;\n }\n kw_args = PyDict_Size(__pyx_kwds);\n switch (pos_args) {\n case 0:\n if (likely((values[0] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_dt)) != 0)) kw_args--;\n else goto __pyx_L5_argtuple_error;\n CYTHON_FALLTHROUGH;\n case 1:\n if (likely((values[1] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_gt)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"_rleIou\", 1, 6, 6, 1); __PYX_ERR(0, 189, __pyx_L3_error)\n }\n CYTHON_FALLTHROUGH;\n case 2:\n if (likely((values[2] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_iscrowd)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"_rleIou\", 1, 6, 6, 2); __PYX_ERR(0, 189, __pyx_L3_error)\n }\n CYTHON_FALLTHROUGH;\n case 3:\n if (likely((values[3] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_m)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"_rleIou\", 1, 6, 6, 3); __PYX_ERR(0, 189, __pyx_L3_error)\n }\n CYTHON_FALLTHROUGH;\n case 4:\n if (likely((values[4] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_n)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"_rleIou\", 1, 6, 6, 4); __PYX_ERR(0, 189, __pyx_L3_error)\n }\n CYTHON_FALLTHROUGH;\n case 5:\n if (likely((values[5] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_iou)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"_rleIou\", 1, 6, 6, 5); __PYX_ERR(0, 189, __pyx_L3_error)\n }\n }\n if (unlikely(kw_args > 0)) {\n if (unlikely(__Pyx_ParseOptionalKeywords(__pyx_kwds, __pyx_pyargnames, 0, values, pos_args, \"_rleIou\") < 0)) __PYX_ERR(0, 189, __pyx_L3_error)\n }\n } else if (PyTuple_GET_SIZE(__pyx_args) != 6) {\n goto __pyx_L5_argtuple_error;\n } else {\n values[0] = PyTuple_GET_ITEM(__pyx_args, 0);\n values[1] = PyTuple_GET_ITEM(__pyx_args, 1);\n values[2] = PyTuple_GET_ITEM(__pyx_args, 2);\n values[3] = PyTuple_GET_ITEM(__pyx_args, 3);\n values[4] = PyTuple_GET_ITEM(__pyx_args, 4);\n values[5] = PyTuple_GET_ITEM(__pyx_args, 5);\n }\n __pyx_v_dt = ((struct __pyx_obj_11pycocotools_5_mask_RLEs *)values[0]);\n __pyx_v_gt = ((struct __pyx_obj_11pycocotools_5_mask_RLEs *)values[1]);\n __pyx_v_iscrowd = ((PyArrayObject *)values[2]);\n __pyx_v_m = __Pyx_PyInt_As_siz(values[3]); if (unlikely((__pyx_v_m == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 189, __pyx_L3_error)\n __pyx_v_n = __Pyx_PyInt_As_siz(values[4]); if (unlikely((__pyx_v_n == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 189, __pyx_L3_error)\n __pyx_v__iou = ((PyArrayObject *)values[5]);\n }\n goto __pyx_L4_argument_unpacking_done;\n __pyx_L5_argtuple_error:;\n __Pyx_RaiseArgtupleInvalid(\"_rleIou\", 1, 6, 6, PyTuple_GET_SIZE(__pyx_args)); __PYX_ERR(0, 189, __pyx_L3_error)\n __pyx_L3_error:;\n __Pyx_AddTraceback(\"pycocotools._mask.iou._rleIou\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __Pyx_RefNannyFinishContext();\n return NULL;\n __pyx_L4_argument_unpacking_done:;\n if (unlikely(!__Pyx_ArgTypeTest(((PyObject *)__pyx_v_dt), __pyx_ptype_11pycocotools_5_mask_RLEs, 1, \"dt\", 0))) __PYX_ERR(0, 189, __pyx_L1_error)\n if (unlikely(!__Pyx_ArgTypeTest(((PyObject *)__pyx_v_gt), __pyx_ptype_11pycocotools_5_mask_RLEs, 1, \"gt\", 0))) __PYX_ERR(0, 189, __pyx_L1_error)\n if (unlikely(!__Pyx_ArgTypeTest(((PyObject *)__pyx_v_iscrowd), __pyx_ptype_5numpy_ndarray, 1, \"iscrowd\", 0))) __PYX_ERR(0, 189, __pyx_L1_error)\n if (unlikely(!__Pyx_ArgTypeTest(((PyObject *)__pyx_v__iou), __pyx_ptype_5numpy_ndarray, 1, \"_iou\", 0))) __PYX_ERR(0, 189, __pyx_L1_error)\n __pyx_r = __pyx_pf_11pycocotools_5_mask_3iou_2_rleIou(__pyx_self, __pyx_v_dt, __pyx_v_gt, __pyx_v_iscrowd, __pyx_v_m, __pyx_v_n, __pyx_v__iou);\n\n /* function exit code */\n goto __pyx_L0;\n __pyx_L1_error:;\n __pyx_r = NULL;\n __pyx_L0:;\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_3iou_2_rleIou(CYTHON_UNUSED PyObject *__pyx_self, struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_dt, struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_gt, PyArrayObject *__pyx_v_iscrowd, siz __pyx_v_m, siz __pyx_v_n, PyArrayObject *__pyx_v__iou) {\n __Pyx_LocalBuf_ND __pyx_pybuffernd__iou;\n __Pyx_Buffer __pyx_pybuffer__iou;\n __Pyx_LocalBuf_ND __pyx_pybuffernd_iscrowd;\n __Pyx_Buffer __pyx_pybuffer_iscrowd;\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"_rleIou\", 0);\n __pyx_pybuffer_iscrowd.pybuffer.buf = NULL;\n __pyx_pybuffer_iscrowd.refcount = 0;\n __pyx_pybuffernd_iscrowd.data = NULL;\n __pyx_pybuffernd_iscrowd.rcbuffer = &__pyx_pybuffer_iscrowd;\n __pyx_pybuffer__iou.pybuffer.buf = NULL;\n __pyx_pybuffer__iou.refcount = 0;\n __pyx_pybuffernd__iou.data = NULL;\n __pyx_pybuffernd__iou.rcbuffer = &__pyx_pybuffer__iou;\n {\n __Pyx_BufFmt_StackElem __pyx_stack[1];\n if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_iscrowd.rcbuffer->pybuffer, (PyObject*)__pyx_v_iscrowd, &__Pyx_TypeInfo_nn___pyx_t_5numpy_uint8_t, PyBUF_FORMAT| PyBUF_STRIDES, 1, 0, __pyx_stack) == -1)) __PYX_ERR(0, 189, __pyx_L1_error)\n }\n __pyx_pybuffernd_iscrowd.diminfo[0].strides = __pyx_pybuffernd_iscrowd.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_iscrowd.diminfo[0].shape = __pyx_pybuffernd_iscrowd.rcbuffer->pybuffer.shape[0];\n {\n __Pyx_BufFmt_StackElem __pyx_stack[1];\n if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd__iou.rcbuffer->pybuffer, (PyObject*)__pyx_v__iou, &__Pyx_TypeInfo_nn___pyx_t_5numpy_double_t, PyBUF_FORMAT| PyBUF_STRIDES, 1, 0, __pyx_stack) == -1)) __PYX_ERR(0, 189, __pyx_L1_error)\n }\n __pyx_pybuffernd__iou.diminfo[0].strides = __pyx_pybuffernd__iou.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd__iou.diminfo[0].shape = __pyx_pybuffernd__iou.rcbuffer->pybuffer.shape[0];\n\n /* \"pycocotools/_mask.pyx\":190\n * return objs\n * def _rleIou(RLEs dt, RLEs gt, np.ndarray[np.uint8_t, ndim=1] iscrowd, siz m, siz n, np.ndarray[np.double_t, ndim=1] _iou):\n * rleIou( dt._R, gt._R, m, n, iscrowd.data, _iou.data ) # <<<<<<<<<<<<<<\n * def _bbIou(np.ndarray[np.double_t, ndim=2] dt, np.ndarray[np.double_t, ndim=2] gt, np.ndarray[np.uint8_t, ndim=1] iscrowd, siz m, siz n, np.ndarray[np.double_t, ndim=1] _iou):\n * bbIou( dt.data, gt.data, m, n, iscrowd.data, _iou.data )\n */\n rleIou(((RLE *)__pyx_v_dt->_R), ((RLE *)__pyx_v_gt->_R), __pyx_v_m, __pyx_v_n, ((byte *)__pyx_v_iscrowd->data), ((double *)__pyx_v__iou->data));\n\n /* \"pycocotools/_mask.pyx\":189\n * raise Exception('unrecognized type. The following type: RLEs (rle), np.ndarray (box), and list (box) are supported.')\n * return objs\n * def _rleIou(RLEs dt, RLEs gt, np.ndarray[np.uint8_t, ndim=1] iscrowd, siz m, siz n, np.ndarray[np.double_t, ndim=1] _iou): # <<<<<<<<<<<<<<\n * rleIou( dt._R, gt._R, m, n, iscrowd.data, _iou.data )\n * def _bbIou(np.ndarray[np.double_t, ndim=2] dt, np.ndarray[np.double_t, ndim=2] gt, np.ndarray[np.uint8_t, ndim=1] iscrowd, siz m, siz n, np.ndarray[np.double_t, ndim=1] _iou):\n */\n\n /* function exit code */\n __pyx_r = Py_None; __Pyx_INCREF(Py_None);\n goto __pyx_L0;\n __pyx_L1_error:;\n { PyObject *__pyx_type, *__pyx_value, *__pyx_tb;\n __Pyx_PyThreadState_declare\n __Pyx_PyThreadState_assign\n __Pyx_ErrFetch(&__pyx_type, &__pyx_value, &__pyx_tb);\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd__iou.rcbuffer->pybuffer);\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_iscrowd.rcbuffer->pybuffer);\n __Pyx_ErrRestore(__pyx_type, __pyx_value, __pyx_tb);}\n __Pyx_AddTraceback(\"pycocotools._mask.iou._rleIou\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n goto __pyx_L2;\n __pyx_L0:;\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd__iou.rcbuffer->pybuffer);\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_iscrowd.rcbuffer->pybuffer);\n __pyx_L2:;\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"pycocotools/_mask.pyx\":191\n * def _rleIou(RLEs dt, RLEs gt, np.ndarray[np.uint8_t, ndim=1] iscrowd, siz m, siz n, np.ndarray[np.double_t, ndim=1] _iou):\n * rleIou( dt._R, gt._R, m, n, iscrowd.data, _iou.data )\n * def _bbIou(np.ndarray[np.double_t, ndim=2] dt, np.ndarray[np.double_t, ndim=2] gt, np.ndarray[np.uint8_t, ndim=1] iscrowd, siz m, siz n, np.ndarray[np.double_t, ndim=1] _iou): # <<<<<<<<<<<<<<\n * bbIou( dt.data, gt.data, m, n, iscrowd.data, _iou.data )\n * def _len(obj):\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_3iou_5_bbIou(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds); /*proto*/\nstatic PyMethodDef __pyx_mdef_11pycocotools_5_mask_3iou_5_bbIou = {\"_bbIou\", (PyCFunction)(void*)(PyCFunctionWithKeywords)__pyx_pw_11pycocotools_5_mask_3iou_5_bbIou, METH_VARARGS|METH_KEYWORDS, 0};\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_3iou_5_bbIou(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds) {\n PyArrayObject *__pyx_v_dt = 0;\n PyArrayObject *__pyx_v_gt = 0;\n PyArrayObject *__pyx_v_iscrowd = 0;\n siz __pyx_v_m;\n siz __pyx_v_n;\n PyArrayObject *__pyx_v__iou = 0;\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"_bbIou (wrapper)\", 0);\n {\n static PyObject **__pyx_pyargnames[] = {&__pyx_n_s_dt,&__pyx_n_s_gt,&__pyx_n_s_iscrowd,&__pyx_n_s_m,&__pyx_n_s_n,&__pyx_n_s_iou,0};\n PyObject* values[6] = {0,0,0,0,0,0};\n if (unlikely(__pyx_kwds)) {\n Py_ssize_t kw_args;\n const Py_ssize_t pos_args = PyTuple_GET_SIZE(__pyx_args);\n switch (pos_args) {\n case 6: values[5] = PyTuple_GET_ITEM(__pyx_args, 5);\n CYTHON_FALLTHROUGH;\n case 5: values[4] = PyTuple_GET_ITEM(__pyx_args, 4);\n CYTHON_FALLTHROUGH;\n case 4: values[3] = PyTuple_GET_ITEM(__pyx_args, 3);\n CYTHON_FALLTHROUGH;\n case 3: values[2] = PyTuple_GET_ITEM(__pyx_args, 2);\n CYTHON_FALLTHROUGH;\n case 2: values[1] = PyTuple_GET_ITEM(__pyx_args, 1);\n CYTHON_FALLTHROUGH;\n case 1: values[0] = PyTuple_GET_ITEM(__pyx_args, 0);\n CYTHON_FALLTHROUGH;\n case 0: break;\n default: goto __pyx_L5_argtuple_error;\n }\n kw_args = PyDict_Size(__pyx_kwds);\n switch (pos_args) {\n case 0:\n if (likely((values[0] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_dt)) != 0)) kw_args--;\n else goto __pyx_L5_argtuple_error;\n CYTHON_FALLTHROUGH;\n case 1:\n if (likely((values[1] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_gt)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"_bbIou\", 1, 6, 6, 1); __PYX_ERR(0, 191, __pyx_L3_error)\n }\n CYTHON_FALLTHROUGH;\n case 2:\n if (likely((values[2] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_iscrowd)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"_bbIou\", 1, 6, 6, 2); __PYX_ERR(0, 191, __pyx_L3_error)\n }\n CYTHON_FALLTHROUGH;\n case 3:\n if (likely((values[3] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_m)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"_bbIou\", 1, 6, 6, 3); __PYX_ERR(0, 191, __pyx_L3_error)\n }\n CYTHON_FALLTHROUGH;\n case 4:\n if (likely((values[4] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_n)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"_bbIou\", 1, 6, 6, 4); __PYX_ERR(0, 191, __pyx_L3_error)\n }\n CYTHON_FALLTHROUGH;\n case 5:\n if (likely((values[5] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_iou)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"_bbIou\", 1, 6, 6, 5); __PYX_ERR(0, 191, __pyx_L3_error)\n }\n }\n if (unlikely(kw_args > 0)) {\n if (unlikely(__Pyx_ParseOptionalKeywords(__pyx_kwds, __pyx_pyargnames, 0, values, pos_args, \"_bbIou\") < 0)) __PYX_ERR(0, 191, __pyx_L3_error)\n }\n } else if (PyTuple_GET_SIZE(__pyx_args) != 6) {\n goto __pyx_L5_argtuple_error;\n } else {\n values[0] = PyTuple_GET_ITEM(__pyx_args, 0);\n values[1] = PyTuple_GET_ITEM(__pyx_args, 1);\n values[2] = PyTuple_GET_ITEM(__pyx_args, 2);\n values[3] = PyTuple_GET_ITEM(__pyx_args, 3);\n values[4] = PyTuple_GET_ITEM(__pyx_args, 4);\n values[5] = PyTuple_GET_ITEM(__pyx_args, 5);\n }\n __pyx_v_dt = ((PyArrayObject *)values[0]);\n __pyx_v_gt = ((PyArrayObject *)values[1]);\n __pyx_v_iscrowd = ((PyArrayObject *)values[2]);\n __pyx_v_m = __Pyx_PyInt_As_siz(values[3]); if (unlikely((__pyx_v_m == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 191, __pyx_L3_error)\n __pyx_v_n = __Pyx_PyInt_As_siz(values[4]); if (unlikely((__pyx_v_n == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 191, __pyx_L3_error)\n __pyx_v__iou = ((PyArrayObject *)values[5]);\n }\n goto __pyx_L4_argument_unpacking_done;\n __pyx_L5_argtuple_error:;\n __Pyx_RaiseArgtupleInvalid(\"_bbIou\", 1, 6, 6, PyTuple_GET_SIZE(__pyx_args)); __PYX_ERR(0, 191, __pyx_L3_error)\n __pyx_L3_error:;\n __Pyx_AddTraceback(\"pycocotools._mask.iou._bbIou\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __Pyx_RefNannyFinishContext();\n return NULL;\n __pyx_L4_argument_unpacking_done:;\n if (unlikely(!__Pyx_ArgTypeTest(((PyObject *)__pyx_v_dt), __pyx_ptype_5numpy_ndarray, 1, \"dt\", 0))) __PYX_ERR(0, 191, __pyx_L1_error)\n if (unlikely(!__Pyx_ArgTypeTest(((PyObject *)__pyx_v_gt), __pyx_ptype_5numpy_ndarray, 1, \"gt\", 0))) __PYX_ERR(0, 191, __pyx_L1_error)\n if (unlikely(!__Pyx_ArgTypeTest(((PyObject *)__pyx_v_iscrowd), __pyx_ptype_5numpy_ndarray, 1, \"iscrowd\", 0))) __PYX_ERR(0, 191, __pyx_L1_error)\n if (unlikely(!__Pyx_ArgTypeTest(((PyObject *)__pyx_v__iou), __pyx_ptype_5numpy_ndarray, 1, \"_iou\", 0))) __PYX_ERR(0, 191, __pyx_L1_error)\n __pyx_r = __pyx_pf_11pycocotools_5_mask_3iou_4_bbIou(__pyx_self, __pyx_v_dt, __pyx_v_gt, __pyx_v_iscrowd, __pyx_v_m, __pyx_v_n, __pyx_v__iou);\n\n /* function exit code */\n goto __pyx_L0;\n __pyx_L1_error:;\n __pyx_r = NULL;\n __pyx_L0:;\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_3iou_4_bbIou(CYTHON_UNUSED PyObject *__pyx_self, PyArrayObject *__pyx_v_dt, PyArrayObject *__pyx_v_gt, PyArrayObject *__pyx_v_iscrowd, siz __pyx_v_m, siz __pyx_v_n, PyArrayObject *__pyx_v__iou) {\n __Pyx_LocalBuf_ND __pyx_pybuffernd__iou;\n __Pyx_Buffer __pyx_pybuffer__iou;\n __Pyx_LocalBuf_ND __pyx_pybuffernd_dt;\n __Pyx_Buffer __pyx_pybuffer_dt;\n __Pyx_LocalBuf_ND __pyx_pybuffernd_gt;\n __Pyx_Buffer __pyx_pybuffer_gt;\n __Pyx_LocalBuf_ND __pyx_pybuffernd_iscrowd;\n __Pyx_Buffer __pyx_pybuffer_iscrowd;\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"_bbIou\", 0);\n __pyx_pybuffer_dt.pybuffer.buf = NULL;\n __pyx_pybuffer_dt.refcount = 0;\n __pyx_pybuffernd_dt.data = NULL;\n __pyx_pybuffernd_dt.rcbuffer = &__pyx_pybuffer_dt;\n __pyx_pybuffer_gt.pybuffer.buf = NULL;\n __pyx_pybuffer_gt.refcount = 0;\n __pyx_pybuffernd_gt.data = NULL;\n __pyx_pybuffernd_gt.rcbuffer = &__pyx_pybuffer_gt;\n __pyx_pybuffer_iscrowd.pybuffer.buf = NULL;\n __pyx_pybuffer_iscrowd.refcount = 0;\n __pyx_pybuffernd_iscrowd.data = NULL;\n __pyx_pybuffernd_iscrowd.rcbuffer = &__pyx_pybuffer_iscrowd;\n __pyx_pybuffer__iou.pybuffer.buf = NULL;\n __pyx_pybuffer__iou.refcount = 0;\n __pyx_pybuffernd__iou.data = NULL;\n __pyx_pybuffernd__iou.rcbuffer = &__pyx_pybuffer__iou;\n {\n __Pyx_BufFmt_StackElem __pyx_stack[1];\n if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_dt.rcbuffer->pybuffer, (PyObject*)__pyx_v_dt, &__Pyx_TypeInfo_nn___pyx_t_5numpy_double_t, PyBUF_FORMAT| PyBUF_STRIDES, 2, 0, __pyx_stack) == -1)) __PYX_ERR(0, 191, __pyx_L1_error)\n }\n __pyx_pybuffernd_dt.diminfo[0].strides = __pyx_pybuffernd_dt.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_dt.diminfo[0].shape = __pyx_pybuffernd_dt.rcbuffer->pybuffer.shape[0]; __pyx_pybuffernd_dt.diminfo[1].strides = __pyx_pybuffernd_dt.rcbuffer->pybuffer.strides[1]; __pyx_pybuffernd_dt.diminfo[1].shape = __pyx_pybuffernd_dt.rcbuffer->pybuffer.shape[1];\n {\n __Pyx_BufFmt_StackElem __pyx_stack[1];\n if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_gt.rcbuffer->pybuffer, (PyObject*)__pyx_v_gt, &__Pyx_TypeInfo_nn___pyx_t_5numpy_double_t, PyBUF_FORMAT| PyBUF_STRIDES, 2, 0, __pyx_stack) == -1)) __PYX_ERR(0, 191, __pyx_L1_error)\n }\n __pyx_pybuffernd_gt.diminfo[0].strides = __pyx_pybuffernd_gt.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_gt.diminfo[0].shape = __pyx_pybuffernd_gt.rcbuffer->pybuffer.shape[0]; __pyx_pybuffernd_gt.diminfo[1].strides = __pyx_pybuffernd_gt.rcbuffer->pybuffer.strides[1]; __pyx_pybuffernd_gt.diminfo[1].shape = __pyx_pybuffernd_gt.rcbuffer->pybuffer.shape[1];\n {\n __Pyx_BufFmt_StackElem __pyx_stack[1];\n if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_iscrowd.rcbuffer->pybuffer, (PyObject*)__pyx_v_iscrowd, &__Pyx_TypeInfo_nn___pyx_t_5numpy_uint8_t, PyBUF_FORMAT| PyBUF_STRIDES, 1, 0, __pyx_stack) == -1)) __PYX_ERR(0, 191, __pyx_L1_error)\n }\n __pyx_pybuffernd_iscrowd.diminfo[0].strides = __pyx_pybuffernd_iscrowd.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_iscrowd.diminfo[0].shape = __pyx_pybuffernd_iscrowd.rcbuffer->pybuffer.shape[0];\n {\n __Pyx_BufFmt_StackElem __pyx_stack[1];\n if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd__iou.rcbuffer->pybuffer, (PyObject*)__pyx_v__iou, &__Pyx_TypeInfo_nn___pyx_t_5numpy_double_t, PyBUF_FORMAT| PyBUF_STRIDES, 1, 0, __pyx_stack) == -1)) __PYX_ERR(0, 191, __pyx_L1_error)\n }\n __pyx_pybuffernd__iou.diminfo[0].strides = __pyx_pybuffernd__iou.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd__iou.diminfo[0].shape = __pyx_pybuffernd__iou.rcbuffer->pybuffer.shape[0];\n\n /* \"pycocotools/_mask.pyx\":192\n * rleIou( dt._R, gt._R, m, n, iscrowd.data, _iou.data )\n * def _bbIou(np.ndarray[np.double_t, ndim=2] dt, np.ndarray[np.double_t, ndim=2] gt, np.ndarray[np.uint8_t, ndim=1] iscrowd, siz m, siz n, np.ndarray[np.double_t, ndim=1] _iou):\n * bbIou( dt.data, gt.data, m, n, iscrowd.data, _iou.data ) # <<<<<<<<<<<<<<\n * def _len(obj):\n * cdef siz N = 0\n */\n bbIou(((BB)__pyx_v_dt->data), ((BB)__pyx_v_gt->data), __pyx_v_m, __pyx_v_n, ((byte *)__pyx_v_iscrowd->data), ((double *)__pyx_v__iou->data));\n\n /* \"pycocotools/_mask.pyx\":191\n * def _rleIou(RLEs dt, RLEs gt, np.ndarray[np.uint8_t, ndim=1] iscrowd, siz m, siz n, np.ndarray[np.double_t, ndim=1] _iou):\n * rleIou( dt._R, gt._R, m, n, iscrowd.data, _iou.data )\n * def _bbIou(np.ndarray[np.double_t, ndim=2] dt, np.ndarray[np.double_t, ndim=2] gt, np.ndarray[np.uint8_t, ndim=1] iscrowd, siz m, siz n, np.ndarray[np.double_t, ndim=1] _iou): # <<<<<<<<<<<<<<\n * bbIou( dt.data, gt.data, m, n, iscrowd.data, _iou.data )\n * def _len(obj):\n */\n\n /* function exit code */\n __pyx_r = Py_None; __Pyx_INCREF(Py_None);\n goto __pyx_L0;\n __pyx_L1_error:;\n { PyObject *__pyx_type, *__pyx_value, *__pyx_tb;\n __Pyx_PyThreadState_declare\n __Pyx_PyThreadState_assign\n __Pyx_ErrFetch(&__pyx_type, &__pyx_value, &__pyx_tb);\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd__iou.rcbuffer->pybuffer);\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_dt.rcbuffer->pybuffer);\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_gt.rcbuffer->pybuffer);\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_iscrowd.rcbuffer->pybuffer);\n __Pyx_ErrRestore(__pyx_type, __pyx_value, __pyx_tb);}\n __Pyx_AddTraceback(\"pycocotools._mask.iou._bbIou\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n goto __pyx_L2;\n __pyx_L0:;\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd__iou.rcbuffer->pybuffer);\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_dt.rcbuffer->pybuffer);\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_gt.rcbuffer->pybuffer);\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_iscrowd.rcbuffer->pybuffer);\n __pyx_L2:;\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"pycocotools/_mask.pyx\":193\n * def _bbIou(np.ndarray[np.double_t, ndim=2] dt, np.ndarray[np.double_t, ndim=2] gt, np.ndarray[np.uint8_t, ndim=1] iscrowd, siz m, siz n, np.ndarray[np.double_t, ndim=1] _iou):\n * bbIou( dt.data, gt.data, m, n, iscrowd.data, _iou.data )\n * def _len(obj): # <<<<<<<<<<<<<<\n * cdef siz N = 0\n * if type(obj) == RLEs:\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_3iou_7_len(PyObject *__pyx_self, PyObject *__pyx_v_obj); /*proto*/\nstatic PyMethodDef __pyx_mdef_11pycocotools_5_mask_3iou_7_len = {\"_len\", (PyCFunction)__pyx_pw_11pycocotools_5_mask_3iou_7_len, METH_O, 0};\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_3iou_7_len(PyObject *__pyx_self, PyObject *__pyx_v_obj) {\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"_len (wrapper)\", 0);\n __pyx_r = __pyx_pf_11pycocotools_5_mask_3iou_6_len(__pyx_self, ((PyObject *)__pyx_v_obj));\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_3iou_6_len(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_obj) {\n siz __pyx_v_N;\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n PyObject *__pyx_t_1 = NULL;\n int __pyx_t_2;\n siz __pyx_t_3;\n Py_ssize_t __pyx_t_4;\n PyObject *__pyx_t_5 = NULL;\n __Pyx_RefNannySetupContext(\"_len\", 0);\n\n /* \"pycocotools/_mask.pyx\":194\n * bbIou( dt.data, gt.data, m, n, iscrowd.data, _iou.data )\n * def _len(obj):\n * cdef siz N = 0 # <<<<<<<<<<<<<<\n * if type(obj) == RLEs:\n * N = obj.n\n */\n __pyx_v_N = 0;\n\n /* \"pycocotools/_mask.pyx\":195\n * def _len(obj):\n * cdef siz N = 0\n * if type(obj) == RLEs: # <<<<<<<<<<<<<<\n * N = obj.n\n * elif len(obj)==0:\n */\n __pyx_t_1 = PyObject_RichCompare(((PyObject *)Py_TYPE(__pyx_v_obj)), ((PyObject *)__pyx_ptype_11pycocotools_5_mask_RLEs), Py_EQ); __Pyx_XGOTREF(__pyx_t_1); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 195, __pyx_L1_error)\n __pyx_t_2 = __Pyx_PyObject_IsTrue(__pyx_t_1); if (unlikely(__pyx_t_2 < 0)) __PYX_ERR(0, 195, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n if (__pyx_t_2) {\n\n /* \"pycocotools/_mask.pyx\":196\n * cdef siz N = 0\n * if type(obj) == RLEs:\n * N = obj.n # <<<<<<<<<<<<<<\n * elif len(obj)==0:\n * pass\n */\n __pyx_t_1 = __Pyx_PyObject_GetAttrStr(__pyx_v_obj, __pyx_n_s_n); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 196, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_3 = __Pyx_PyInt_As_siz(__pyx_t_1); if (unlikely((__pyx_t_3 == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 196, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __pyx_v_N = __pyx_t_3;\n\n /* \"pycocotools/_mask.pyx\":195\n * def _len(obj):\n * cdef siz N = 0\n * if type(obj) == RLEs: # <<<<<<<<<<<<<<\n * N = obj.n\n * elif len(obj)==0:\n */\n goto __pyx_L3;\n }\n\n /* \"pycocotools/_mask.pyx\":197\n * if type(obj) == RLEs:\n * N = obj.n\n * elif len(obj)==0: # <<<<<<<<<<<<<<\n * pass\n * elif type(obj) == np.ndarray:\n */\n __pyx_t_4 = PyObject_Length(__pyx_v_obj); if (unlikely(__pyx_t_4 == ((Py_ssize_t)-1))) __PYX_ERR(0, 197, __pyx_L1_error)\n __pyx_t_2 = ((__pyx_t_4 == 0) != 0);\n if (__pyx_t_2) {\n goto __pyx_L3;\n }\n\n /* \"pycocotools/_mask.pyx\":199\n * elif len(obj)==0:\n * pass\n * elif type(obj) == np.ndarray: # <<<<<<<<<<<<<<\n * N = obj.shape[0]\n * return N\n */\n __pyx_t_1 = PyObject_RichCompare(((PyObject *)Py_TYPE(__pyx_v_obj)), ((PyObject *)__pyx_ptype_5numpy_ndarray), Py_EQ); __Pyx_XGOTREF(__pyx_t_1); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 199, __pyx_L1_error)\n __pyx_t_2 = __Pyx_PyObject_IsTrue(__pyx_t_1); if (unlikely(__pyx_t_2 < 0)) __PYX_ERR(0, 199, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n if (__pyx_t_2) {\n\n /* \"pycocotools/_mask.pyx\":200\n * pass\n * elif type(obj) == np.ndarray:\n * N = obj.shape[0] # <<<<<<<<<<<<<<\n * return N\n * # convert iscrowd to numpy array\n */\n __pyx_t_1 = __Pyx_PyObject_GetAttrStr(__pyx_v_obj, __pyx_n_s_shape); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 200, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_5 = __Pyx_GetItemInt(__pyx_t_1, 0, long, 1, __Pyx_PyInt_From_long, 0, 0, 1); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 200, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __pyx_t_3 = __Pyx_PyInt_As_siz(__pyx_t_5); if (unlikely((__pyx_t_3 == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 200, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __pyx_v_N = __pyx_t_3;\n\n /* \"pycocotools/_mask.pyx\":199\n * elif len(obj)==0:\n * pass\n * elif type(obj) == np.ndarray: # <<<<<<<<<<<<<<\n * N = obj.shape[0]\n * return N\n */\n }\n __pyx_L3:;\n\n /* \"pycocotools/_mask.pyx\":201\n * elif type(obj) == np.ndarray:\n * N = obj.shape[0]\n * return N # <<<<<<<<<<<<<<\n * # convert iscrowd to numpy array\n * cdef np.ndarray[np.uint8_t, ndim=1] iscrowd = np.array(pyiscrowd, dtype=np.uint8)\n */\n __Pyx_XDECREF(__pyx_r);\n __pyx_t_5 = __Pyx_PyInt_From_siz(__pyx_v_N); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 201, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __pyx_r = __pyx_t_5;\n __pyx_t_5 = 0;\n goto __pyx_L0;\n\n /* \"pycocotools/_mask.pyx\":193\n * def _bbIou(np.ndarray[np.double_t, ndim=2] dt, np.ndarray[np.double_t, ndim=2] gt, np.ndarray[np.uint8_t, ndim=1] iscrowd, siz m, siz n, np.ndarray[np.double_t, ndim=1] _iou):\n * bbIou( dt.data, gt.data, m, n, iscrowd.data, _iou.data )\n * def _len(obj): # <<<<<<<<<<<<<<\n * cdef siz N = 0\n * if type(obj) == RLEs:\n */\n\n /* function exit code */\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_1);\n __Pyx_XDECREF(__pyx_t_5);\n __Pyx_AddTraceback(\"pycocotools._mask.iou._len\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n __pyx_L0:;\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"pycocotools/_mask.pyx\":163\n * \n * # iou computation. support function overload (RLEs-RLEs and bbox-bbox).\n * def iou( dt, gt, pyiscrowd ): # <<<<<<<<<<<<<<\n * def _preproc(objs):\n * if len(objs) == 0:\n */\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_12iou(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_dt, PyObject *__pyx_v_gt, PyObject *__pyx_v_pyiscrowd) {\n PyObject *__pyx_v__preproc = 0;\n PyObject *__pyx_v__rleIou = 0;\n PyObject *__pyx_v__bbIou = 0;\n PyObject *__pyx_v__len = 0;\n PyArrayObject *__pyx_v_iscrowd = 0;\n siz __pyx_v_m;\n siz __pyx_v_n;\n double *__pyx_v__iou;\n npy_intp __pyx_v_shape[1];\n PyObject *__pyx_v__iouFun = NULL;\n PyObject *__pyx_v_iou = NULL;\n __Pyx_LocalBuf_ND __pyx_pybuffernd_iscrowd;\n __Pyx_Buffer __pyx_pybuffer_iscrowd;\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n PyObject *__pyx_t_1 = NULL;\n PyObject *__pyx_t_2 = NULL;\n PyObject *__pyx_t_3 = NULL;\n PyObject *__pyx_t_4 = NULL;\n PyObject *__pyx_t_5 = NULL;\n PyArrayObject *__pyx_t_6 = NULL;\n siz __pyx_t_7;\n int __pyx_t_8;\n int __pyx_t_9;\n int __pyx_t_10;\n PyObject *__pyx_t_11 = NULL;\n __Pyx_RefNannySetupContext(\"iou\", 0);\n __Pyx_INCREF(__pyx_v_dt);\n __Pyx_INCREF(__pyx_v_gt);\n __pyx_pybuffer_iscrowd.pybuffer.buf = NULL;\n __pyx_pybuffer_iscrowd.refcount = 0;\n __pyx_pybuffernd_iscrowd.data = NULL;\n __pyx_pybuffernd_iscrowd.rcbuffer = &__pyx_pybuffer_iscrowd;\n\n /* \"pycocotools/_mask.pyx\":164\n * # iou computation. support function overload (RLEs-RLEs and bbox-bbox).\n * def iou( dt, gt, pyiscrowd ):\n * def _preproc(objs): # <<<<<<<<<<<<<<\n * if len(objs) == 0:\n * return objs\n */\n __pyx_t_1 = __Pyx_CyFunction_NewEx(&__pyx_mdef_11pycocotools_5_mask_3iou_1_preproc, 0, __pyx_n_s_iou_locals__preproc, NULL, __pyx_n_s_pycocotools__mask, __pyx_d, ((PyObject *)__pyx_codeobj__10)); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 164, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_v__preproc = __pyx_t_1;\n __pyx_t_1 = 0;\n\n /* \"pycocotools/_mask.pyx\":189\n * raise Exception('unrecognized type. The following type: RLEs (rle), np.ndarray (box), and list (box) are supported.')\n * return objs\n * def _rleIou(RLEs dt, RLEs gt, np.ndarray[np.uint8_t, ndim=1] iscrowd, siz m, siz n, np.ndarray[np.double_t, ndim=1] _iou): # <<<<<<<<<<<<<<\n * rleIou( dt._R, gt._R, m, n, iscrowd.data, _iou.data )\n * def _bbIou(np.ndarray[np.double_t, ndim=2] dt, np.ndarray[np.double_t, ndim=2] gt, np.ndarray[np.uint8_t, ndim=1] iscrowd, siz m, siz n, np.ndarray[np.double_t, ndim=1] _iou):\n */\n __pyx_t_1 = __Pyx_CyFunction_NewEx(&__pyx_mdef_11pycocotools_5_mask_3iou_3_rleIou, 0, __pyx_n_s_iou_locals__rleIou, NULL, __pyx_n_s_pycocotools__mask, __pyx_d, ((PyObject *)__pyx_codeobj__12)); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 189, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_v__rleIou = __pyx_t_1;\n __pyx_t_1 = 0;\n\n /* \"pycocotools/_mask.pyx\":191\n * def _rleIou(RLEs dt, RLEs gt, np.ndarray[np.uint8_t, ndim=1] iscrowd, siz m, siz n, np.ndarray[np.double_t, ndim=1] _iou):\n * rleIou( dt._R, gt._R, m, n, iscrowd.data, _iou.data )\n * def _bbIou(np.ndarray[np.double_t, ndim=2] dt, np.ndarray[np.double_t, ndim=2] gt, np.ndarray[np.uint8_t, ndim=1] iscrowd, siz m, siz n, np.ndarray[np.double_t, ndim=1] _iou): # <<<<<<<<<<<<<<\n * bbIou( dt.data, gt.data, m, n, iscrowd.data, _iou.data )\n * def _len(obj):\n */\n __pyx_t_1 = __Pyx_CyFunction_NewEx(&__pyx_mdef_11pycocotools_5_mask_3iou_5_bbIou, 0, __pyx_n_s_iou_locals__bbIou, NULL, __pyx_n_s_pycocotools__mask, __pyx_d, ((PyObject *)__pyx_codeobj__14)); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 191, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_v__bbIou = __pyx_t_1;\n __pyx_t_1 = 0;\n\n /* \"pycocotools/_mask.pyx\":193\n * def _bbIou(np.ndarray[np.double_t, ndim=2] dt, np.ndarray[np.double_t, ndim=2] gt, np.ndarray[np.uint8_t, ndim=1] iscrowd, siz m, siz n, np.ndarray[np.double_t, ndim=1] _iou):\n * bbIou( dt.data, gt.data, m, n, iscrowd.data, _iou.data )\n * def _len(obj): # <<<<<<<<<<<<<<\n * cdef siz N = 0\n * if type(obj) == RLEs:\n */\n __pyx_t_1 = __Pyx_CyFunction_NewEx(&__pyx_mdef_11pycocotools_5_mask_3iou_7_len, 0, __pyx_n_s_iou_locals__len, NULL, __pyx_n_s_pycocotools__mask, __pyx_d, ((PyObject *)__pyx_codeobj__16)); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 193, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_v__len = __pyx_t_1;\n __pyx_t_1 = 0;\n\n /* \"pycocotools/_mask.pyx\":203\n * return N\n * # convert iscrowd to numpy array\n * cdef np.ndarray[np.uint8_t, ndim=1] iscrowd = np.array(pyiscrowd, dtype=np.uint8) # <<<<<<<<<<<<<<\n * # simple type checking\n * cdef siz m, n\n */\n __Pyx_GetModuleGlobalName(__pyx_t_1, __pyx_n_s_np); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 203, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_2 = __Pyx_PyObject_GetAttrStr(__pyx_t_1, __pyx_n_s_array); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 203, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __pyx_t_1 = PyTuple_New(1); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 203, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_INCREF(__pyx_v_pyiscrowd);\n __Pyx_GIVEREF(__pyx_v_pyiscrowd);\n PyTuple_SET_ITEM(__pyx_t_1, 0, __pyx_v_pyiscrowd);\n __pyx_t_3 = __Pyx_PyDict_NewPresized(1); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 203, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_GetModuleGlobalName(__pyx_t_4, __pyx_n_s_np); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 203, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __pyx_t_5 = __Pyx_PyObject_GetAttrStr(__pyx_t_4, __pyx_n_s_uint8); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 203, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;\n if (PyDict_SetItem(__pyx_t_3, __pyx_n_s_dtype, __pyx_t_5) < 0) __PYX_ERR(0, 203, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __pyx_t_5 = __Pyx_PyObject_Call(__pyx_t_2, __pyx_t_1, __pyx_t_3); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 203, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n if (!(likely(((__pyx_t_5) == Py_None) || likely(__Pyx_TypeTest(__pyx_t_5, __pyx_ptype_5numpy_ndarray))))) __PYX_ERR(0, 203, __pyx_L1_error)\n __pyx_t_6 = ((PyArrayObject *)__pyx_t_5);\n {\n __Pyx_BufFmt_StackElem __pyx_stack[1];\n if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_iscrowd.rcbuffer->pybuffer, (PyObject*)__pyx_t_6, &__Pyx_TypeInfo_nn___pyx_t_5numpy_uint8_t, PyBUF_FORMAT| PyBUF_STRIDES, 1, 0, __pyx_stack) == -1)) {\n __pyx_v_iscrowd = ((PyArrayObject *)Py_None); __Pyx_INCREF(Py_None); __pyx_pybuffernd_iscrowd.rcbuffer->pybuffer.buf = NULL;\n __PYX_ERR(0, 203, __pyx_L1_error)\n } else {__pyx_pybuffernd_iscrowd.diminfo[0].strides = __pyx_pybuffernd_iscrowd.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_iscrowd.diminfo[0].shape = __pyx_pybuffernd_iscrowd.rcbuffer->pybuffer.shape[0];\n }\n }\n __pyx_t_6 = 0;\n __pyx_v_iscrowd = ((PyArrayObject *)__pyx_t_5);\n __pyx_t_5 = 0;\n\n /* \"pycocotools/_mask.pyx\":206\n * # simple type checking\n * cdef siz m, n\n * dt = _preproc(dt) # <<<<<<<<<<<<<<\n * gt = _preproc(gt)\n * m = _len(dt)\n */\n __pyx_t_5 = __pyx_pf_11pycocotools_5_mask_3iou__preproc(__pyx_v__preproc, __pyx_v_dt); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 206, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_DECREF_SET(__pyx_v_dt, __pyx_t_5);\n __pyx_t_5 = 0;\n\n /* \"pycocotools/_mask.pyx\":207\n * cdef siz m, n\n * dt = _preproc(dt)\n * gt = _preproc(gt) # <<<<<<<<<<<<<<\n * m = _len(dt)\n * n = _len(gt)\n */\n __pyx_t_5 = __pyx_pf_11pycocotools_5_mask_3iou__preproc(__pyx_v__preproc, __pyx_v_gt); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 207, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_DECREF_SET(__pyx_v_gt, __pyx_t_5);\n __pyx_t_5 = 0;\n\n /* \"pycocotools/_mask.pyx\":208\n * dt = _preproc(dt)\n * gt = _preproc(gt)\n * m = _len(dt) # <<<<<<<<<<<<<<\n * n = _len(gt)\n * if m == 0 or n == 0:\n */\n __pyx_t_5 = __pyx_pf_11pycocotools_5_mask_3iou_6_len(__pyx_v__len, __pyx_v_dt); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 208, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __pyx_t_7 = __Pyx_PyInt_As_siz(__pyx_t_5); if (unlikely((__pyx_t_7 == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 208, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __pyx_v_m = __pyx_t_7;\n\n /* \"pycocotools/_mask.pyx\":209\n * gt = _preproc(gt)\n * m = _len(dt)\n * n = _len(gt) # <<<<<<<<<<<<<<\n * if m == 0 or n == 0:\n * return []\n */\n __pyx_t_5 = __pyx_pf_11pycocotools_5_mask_3iou_6_len(__pyx_v__len, __pyx_v_gt); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 209, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __pyx_t_7 = __Pyx_PyInt_As_siz(__pyx_t_5); if (unlikely((__pyx_t_7 == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 209, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __pyx_v_n = __pyx_t_7;\n\n /* \"pycocotools/_mask.pyx\":210\n * m = _len(dt)\n * n = _len(gt)\n * if m == 0 or n == 0: # <<<<<<<<<<<<<<\n * return []\n * if not type(dt) == type(gt):\n */\n __pyx_t_9 = ((__pyx_v_m == 0) != 0);\n if (!__pyx_t_9) {\n } else {\n __pyx_t_8 = __pyx_t_9;\n goto __pyx_L4_bool_binop_done;\n }\n __pyx_t_9 = ((__pyx_v_n == 0) != 0);\n __pyx_t_8 = __pyx_t_9;\n __pyx_L4_bool_binop_done:;\n if (__pyx_t_8) {\n\n /* \"pycocotools/_mask.pyx\":211\n * n = _len(gt)\n * if m == 0 or n == 0:\n * return [] # <<<<<<<<<<<<<<\n * if not type(dt) == type(gt):\n * raise Exception('The dt and gt should have the same data type, either RLEs, list or np.ndarray')\n */\n __Pyx_XDECREF(__pyx_r);\n __pyx_t_5 = PyList_New(0); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 211, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __pyx_r = __pyx_t_5;\n __pyx_t_5 = 0;\n goto __pyx_L0;\n\n /* \"pycocotools/_mask.pyx\":210\n * m = _len(dt)\n * n = _len(gt)\n * if m == 0 or n == 0: # <<<<<<<<<<<<<<\n * return []\n * if not type(dt) == type(gt):\n */\n }\n\n /* \"pycocotools/_mask.pyx\":212\n * if m == 0 or n == 0:\n * return []\n * if not type(dt) == type(gt): # <<<<<<<<<<<<<<\n * raise Exception('The dt and gt should have the same data type, either RLEs, list or np.ndarray')\n * \n */\n __pyx_t_5 = PyObject_RichCompare(((PyObject *)Py_TYPE(__pyx_v_dt)), ((PyObject *)Py_TYPE(__pyx_v_gt)), Py_EQ); __Pyx_XGOTREF(__pyx_t_5); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 212, __pyx_L1_error)\n __pyx_t_8 = __Pyx_PyObject_IsTrue(__pyx_t_5); if (unlikely(__pyx_t_8 < 0)) __PYX_ERR(0, 212, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __pyx_t_9 = ((!__pyx_t_8) != 0);\n if (unlikely(__pyx_t_9)) {\n\n /* \"pycocotools/_mask.pyx\":213\n * return []\n * if not type(dt) == type(gt):\n * raise Exception('The dt and gt should have the same data type, either RLEs, list or np.ndarray') # <<<<<<<<<<<<<<\n * \n * # define local variables\n */\n __pyx_t_5 = __Pyx_PyObject_Call(((PyObject *)(&((PyTypeObject*)PyExc_Exception)[0])), __pyx_tuple__17, NULL); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 213, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_Raise(__pyx_t_5, 0, 0, 0);\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __PYX_ERR(0, 213, __pyx_L1_error)\n\n /* \"pycocotools/_mask.pyx\":212\n * if m == 0 or n == 0:\n * return []\n * if not type(dt) == type(gt): # <<<<<<<<<<<<<<\n * raise Exception('The dt and gt should have the same data type, either RLEs, list or np.ndarray')\n * \n */\n }\n\n /* \"pycocotools/_mask.pyx\":216\n * \n * # define local variables\n * cdef double* _iou = 0 # <<<<<<<<<<<<<<\n * cdef np.npy_intp shape[1]\n * # check type and assign iou function\n */\n __pyx_v__iou = ((double *)0);\n\n /* \"pycocotools/_mask.pyx\":219\n * cdef np.npy_intp shape[1]\n * # check type and assign iou function\n * if type(dt) == RLEs: # <<<<<<<<<<<<<<\n * _iouFun = _rleIou\n * elif type(dt) == np.ndarray:\n */\n __pyx_t_5 = PyObject_RichCompare(((PyObject *)Py_TYPE(__pyx_v_dt)), ((PyObject *)__pyx_ptype_11pycocotools_5_mask_RLEs), Py_EQ); __Pyx_XGOTREF(__pyx_t_5); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 219, __pyx_L1_error)\n __pyx_t_9 = __Pyx_PyObject_IsTrue(__pyx_t_5); if (unlikely(__pyx_t_9 < 0)) __PYX_ERR(0, 219, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n if (__pyx_t_9) {\n\n /* \"pycocotools/_mask.pyx\":220\n * # check type and assign iou function\n * if type(dt) == RLEs:\n * _iouFun = _rleIou # <<<<<<<<<<<<<<\n * elif type(dt) == np.ndarray:\n * _iouFun = _bbIou\n */\n __Pyx_INCREF(__pyx_v__rleIou);\n __pyx_v__iouFun = __pyx_v__rleIou;\n\n /* \"pycocotools/_mask.pyx\":219\n * cdef np.npy_intp shape[1]\n * # check type and assign iou function\n * if type(dt) == RLEs: # <<<<<<<<<<<<<<\n * _iouFun = _rleIou\n * elif type(dt) == np.ndarray:\n */\n goto __pyx_L7;\n }\n\n /* \"pycocotools/_mask.pyx\":221\n * if type(dt) == RLEs:\n * _iouFun = _rleIou\n * elif type(dt) == np.ndarray: # <<<<<<<<<<<<<<\n * _iouFun = _bbIou\n * else:\n */\n __pyx_t_5 = PyObject_RichCompare(((PyObject *)Py_TYPE(__pyx_v_dt)), ((PyObject *)__pyx_ptype_5numpy_ndarray), Py_EQ); __Pyx_XGOTREF(__pyx_t_5); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 221, __pyx_L1_error)\n __pyx_t_9 = __Pyx_PyObject_IsTrue(__pyx_t_5); if (unlikely(__pyx_t_9 < 0)) __PYX_ERR(0, 221, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n if (likely(__pyx_t_9)) {\n\n /* \"pycocotools/_mask.pyx\":222\n * _iouFun = _rleIou\n * elif type(dt) == np.ndarray:\n * _iouFun = _bbIou # <<<<<<<<<<<<<<\n * else:\n * raise Exception('input data type not allowed.')\n */\n __Pyx_INCREF(__pyx_v__bbIou);\n __pyx_v__iouFun = __pyx_v__bbIou;\n\n /* \"pycocotools/_mask.pyx\":221\n * if type(dt) == RLEs:\n * _iouFun = _rleIou\n * elif type(dt) == np.ndarray: # <<<<<<<<<<<<<<\n * _iouFun = _bbIou\n * else:\n */\n goto __pyx_L7;\n }\n\n /* \"pycocotools/_mask.pyx\":224\n * _iouFun = _bbIou\n * else:\n * raise Exception('input data type not allowed.') # <<<<<<<<<<<<<<\n * _iou = malloc(m*n* sizeof(double))\n * iou = np.zeros((m*n, ), dtype=np.double)\n */\n /*else*/ {\n __pyx_t_5 = __Pyx_PyObject_Call(((PyObject *)(&((PyTypeObject*)PyExc_Exception)[0])), __pyx_tuple__18, NULL); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 224, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_Raise(__pyx_t_5, 0, 0, 0);\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __PYX_ERR(0, 224, __pyx_L1_error)\n }\n __pyx_L7:;\n\n /* \"pycocotools/_mask.pyx\":225\n * else:\n * raise Exception('input data type not allowed.')\n * _iou = malloc(m*n* sizeof(double)) # <<<<<<<<<<<<<<\n * iou = np.zeros((m*n, ), dtype=np.double)\n * shape[0] = m*n\n */\n __pyx_v__iou = ((double *)malloc(((__pyx_v_m * __pyx_v_n) * (sizeof(double)))));\n\n /* \"pycocotools/_mask.pyx\":226\n * raise Exception('input data type not allowed.')\n * _iou = malloc(m*n* sizeof(double))\n * iou = np.zeros((m*n, ), dtype=np.double) # <<<<<<<<<<<<<<\n * shape[0] = m*n\n * iou = np.PyArray_SimpleNewFromData(1, shape, np.NPY_DOUBLE, _iou)\n */\n __Pyx_GetModuleGlobalName(__pyx_t_5, __pyx_n_s_np); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 226, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __pyx_t_3 = __Pyx_PyObject_GetAttrStr(__pyx_t_5, __pyx_n_s_zeros); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 226, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __pyx_t_5 = __Pyx_PyInt_From_siz((__pyx_v_m * __pyx_v_n)); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 226, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __pyx_t_1 = PyTuple_New(1); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 226, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_GIVEREF(__pyx_t_5);\n PyTuple_SET_ITEM(__pyx_t_1, 0, __pyx_t_5);\n __pyx_t_5 = 0;\n __pyx_t_5 = PyTuple_New(1); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 226, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_GIVEREF(__pyx_t_1);\n PyTuple_SET_ITEM(__pyx_t_5, 0, __pyx_t_1);\n __pyx_t_1 = 0;\n __pyx_t_1 = __Pyx_PyDict_NewPresized(1); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 226, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_GetModuleGlobalName(__pyx_t_2, __pyx_n_s_np); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 226, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __pyx_t_4 = __Pyx_PyObject_GetAttrStr(__pyx_t_2, __pyx_n_s_double); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 226, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n if (PyDict_SetItem(__pyx_t_1, __pyx_n_s_dtype, __pyx_t_4) < 0) __PYX_ERR(0, 226, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;\n __pyx_t_4 = __Pyx_PyObject_Call(__pyx_t_3, __pyx_t_5, __pyx_t_1); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 226, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __pyx_v_iou = __pyx_t_4;\n __pyx_t_4 = 0;\n\n /* \"pycocotools/_mask.pyx\":227\n * _iou = malloc(m*n* sizeof(double))\n * iou = np.zeros((m*n, ), dtype=np.double)\n * shape[0] = m*n # <<<<<<<<<<<<<<\n * iou = np.PyArray_SimpleNewFromData(1, shape, np.NPY_DOUBLE, _iou)\n * PyArray_ENABLEFLAGS(iou, np.NPY_OWNDATA)\n */\n (__pyx_v_shape[0]) = (((npy_intp)__pyx_v_m) * __pyx_v_n);\n\n /* \"pycocotools/_mask.pyx\":228\n * iou = np.zeros((m*n, ), dtype=np.double)\n * shape[0] = m*n\n * iou = np.PyArray_SimpleNewFromData(1, shape, np.NPY_DOUBLE, _iou) # <<<<<<<<<<<<<<\n * PyArray_ENABLEFLAGS(iou, np.NPY_OWNDATA)\n * _iouFun(dt, gt, iscrowd, m, n, iou)\n */\n __pyx_t_4 = PyArray_SimpleNewFromData(1, __pyx_v_shape, NPY_DOUBLE, __pyx_v__iou); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 228, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __Pyx_DECREF_SET(__pyx_v_iou, __pyx_t_4);\n __pyx_t_4 = 0;\n\n /* \"pycocotools/_mask.pyx\":229\n * shape[0] = m*n\n * iou = np.PyArray_SimpleNewFromData(1, shape, np.NPY_DOUBLE, _iou)\n * PyArray_ENABLEFLAGS(iou, np.NPY_OWNDATA) # <<<<<<<<<<<<<<\n * _iouFun(dt, gt, iscrowd, m, n, iou)\n * return iou.reshape((m,n), order='F')\n */\n if (!(likely(((__pyx_v_iou) == Py_None) || likely(__Pyx_TypeTest(__pyx_v_iou, __pyx_ptype_5numpy_ndarray))))) __PYX_ERR(0, 229, __pyx_L1_error)\n PyArray_ENABLEFLAGS(((PyArrayObject *)__pyx_v_iou), NPY_OWNDATA);\n\n /* \"pycocotools/_mask.pyx\":230\n * iou = np.PyArray_SimpleNewFromData(1, shape, np.NPY_DOUBLE, _iou)\n * PyArray_ENABLEFLAGS(iou, np.NPY_OWNDATA)\n * _iouFun(dt, gt, iscrowd, m, n, iou) # <<<<<<<<<<<<<<\n * return iou.reshape((m,n), order='F')\n * \n */\n __pyx_t_1 = __Pyx_PyInt_From_siz(__pyx_v_m); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 230, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_5 = __Pyx_PyInt_From_siz(__pyx_v_n); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 230, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_INCREF(__pyx_v__iouFun);\n __pyx_t_3 = __pyx_v__iouFun; __pyx_t_2 = NULL;\n __pyx_t_10 = 0;\n if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_3))) {\n __pyx_t_2 = PyMethod_GET_SELF(__pyx_t_3);\n if (likely(__pyx_t_2)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_3);\n __Pyx_INCREF(__pyx_t_2);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_3, function);\n __pyx_t_10 = 1;\n }\n }\n #if CYTHON_FAST_PYCALL\n if (PyFunction_Check(__pyx_t_3)) {\n PyObject *__pyx_temp[7] = {__pyx_t_2, __pyx_v_dt, __pyx_v_gt, ((PyObject *)__pyx_v_iscrowd), __pyx_t_1, __pyx_t_5, __pyx_v_iou};\n __pyx_t_4 = __Pyx_PyFunction_FastCall(__pyx_t_3, __pyx_temp+1-__pyx_t_10, 6+__pyx_t_10); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 230, __pyx_L1_error)\n __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;\n __Pyx_GOTREF(__pyx_t_4);\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n } else\n #endif\n #if CYTHON_FAST_PYCCALL\n if (__Pyx_PyFastCFunction_Check(__pyx_t_3)) {\n PyObject *__pyx_temp[7] = {__pyx_t_2, __pyx_v_dt, __pyx_v_gt, ((PyObject *)__pyx_v_iscrowd), __pyx_t_1, __pyx_t_5, __pyx_v_iou};\n __pyx_t_4 = __Pyx_PyCFunction_FastCall(__pyx_t_3, __pyx_temp+1-__pyx_t_10, 6+__pyx_t_10); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 230, __pyx_L1_error)\n __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;\n __Pyx_GOTREF(__pyx_t_4);\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n } else\n #endif\n {\n __pyx_t_11 = PyTuple_New(6+__pyx_t_10); if (unlikely(!__pyx_t_11)) __PYX_ERR(0, 230, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_11);\n if (__pyx_t_2) {\n __Pyx_GIVEREF(__pyx_t_2); PyTuple_SET_ITEM(__pyx_t_11, 0, __pyx_t_2); __pyx_t_2 = NULL;\n }\n __Pyx_INCREF(__pyx_v_dt);\n __Pyx_GIVEREF(__pyx_v_dt);\n PyTuple_SET_ITEM(__pyx_t_11, 0+__pyx_t_10, __pyx_v_dt);\n __Pyx_INCREF(__pyx_v_gt);\n __Pyx_GIVEREF(__pyx_v_gt);\n PyTuple_SET_ITEM(__pyx_t_11, 1+__pyx_t_10, __pyx_v_gt);\n __Pyx_INCREF(((PyObject *)__pyx_v_iscrowd));\n __Pyx_GIVEREF(((PyObject *)__pyx_v_iscrowd));\n PyTuple_SET_ITEM(__pyx_t_11, 2+__pyx_t_10, ((PyObject *)__pyx_v_iscrowd));\n __Pyx_GIVEREF(__pyx_t_1);\n PyTuple_SET_ITEM(__pyx_t_11, 3+__pyx_t_10, __pyx_t_1);\n __Pyx_GIVEREF(__pyx_t_5);\n PyTuple_SET_ITEM(__pyx_t_11, 4+__pyx_t_10, __pyx_t_5);\n __Pyx_INCREF(__pyx_v_iou);\n __Pyx_GIVEREF(__pyx_v_iou);\n PyTuple_SET_ITEM(__pyx_t_11, 5+__pyx_t_10, __pyx_v_iou);\n __pyx_t_1 = 0;\n __pyx_t_5 = 0;\n __pyx_t_4 = __Pyx_PyObject_Call(__pyx_t_3, __pyx_t_11, NULL); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 230, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __Pyx_DECREF(__pyx_t_11); __pyx_t_11 = 0;\n }\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;\n\n /* \"pycocotools/_mask.pyx\":231\n * PyArray_ENABLEFLAGS(iou, np.NPY_OWNDATA)\n * _iouFun(dt, gt, iscrowd, m, n, iou)\n * return iou.reshape((m,n), order='F') # <<<<<<<<<<<<<<\n * \n * def toBbox( rleObjs ):\n */\n __Pyx_XDECREF(__pyx_r);\n __pyx_t_4 = __Pyx_PyObject_GetAttrStr(__pyx_v_iou, __pyx_n_s_reshape); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 231, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __pyx_t_3 = __Pyx_PyInt_From_siz(__pyx_v_m); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 231, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __pyx_t_11 = __Pyx_PyInt_From_siz(__pyx_v_n); if (unlikely(!__pyx_t_11)) __PYX_ERR(0, 231, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_11);\n __pyx_t_5 = PyTuple_New(2); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 231, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_GIVEREF(__pyx_t_3);\n PyTuple_SET_ITEM(__pyx_t_5, 0, __pyx_t_3);\n __Pyx_GIVEREF(__pyx_t_11);\n PyTuple_SET_ITEM(__pyx_t_5, 1, __pyx_t_11);\n __pyx_t_3 = 0;\n __pyx_t_11 = 0;\n __pyx_t_11 = PyTuple_New(1); if (unlikely(!__pyx_t_11)) __PYX_ERR(0, 231, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_11);\n __Pyx_GIVEREF(__pyx_t_5);\n PyTuple_SET_ITEM(__pyx_t_11, 0, __pyx_t_5);\n __pyx_t_5 = 0;\n __pyx_t_5 = __Pyx_PyDict_NewPresized(1); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 231, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n if (PyDict_SetItem(__pyx_t_5, __pyx_n_s_order, __pyx_n_s_F) < 0) __PYX_ERR(0, 231, __pyx_L1_error)\n __pyx_t_3 = __Pyx_PyObject_Call(__pyx_t_4, __pyx_t_11, __pyx_t_5); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 231, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;\n __Pyx_DECREF(__pyx_t_11); __pyx_t_11 = 0;\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __pyx_r = __pyx_t_3;\n __pyx_t_3 = 0;\n goto __pyx_L0;\n\n /* \"pycocotools/_mask.pyx\":163\n * \n * # iou computation. support function overload (RLEs-RLEs and bbox-bbox).\n * def iou( dt, gt, pyiscrowd ): # <<<<<<<<<<<<<<\n * def _preproc(objs):\n * if len(objs) == 0:\n */\n\n /* function exit code */\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_1);\n __Pyx_XDECREF(__pyx_t_2);\n __Pyx_XDECREF(__pyx_t_3);\n __Pyx_XDECREF(__pyx_t_4);\n __Pyx_XDECREF(__pyx_t_5);\n __Pyx_XDECREF(__pyx_t_11);\n { PyObject *__pyx_type, *__pyx_value, *__pyx_tb;\n __Pyx_PyThreadState_declare\n __Pyx_PyThreadState_assign\n __Pyx_ErrFetch(&__pyx_type, &__pyx_value, &__pyx_tb);\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_iscrowd.rcbuffer->pybuffer);\n __Pyx_ErrRestore(__pyx_type, __pyx_value, __pyx_tb);}\n __Pyx_AddTraceback(\"pycocotools._mask.iou\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n goto __pyx_L2;\n __pyx_L0:;\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_iscrowd.rcbuffer->pybuffer);\n __pyx_L2:;\n __Pyx_XDECREF(__pyx_v__preproc);\n __Pyx_XDECREF(__pyx_v__rleIou);\n __Pyx_XDECREF(__pyx_v__bbIou);\n __Pyx_XDECREF(__pyx_v__len);\n __Pyx_XDECREF((PyObject *)__pyx_v_iscrowd);\n __Pyx_XDECREF(__pyx_v__iouFun);\n __Pyx_XDECREF(__pyx_v_iou);\n __Pyx_XDECREF(__pyx_v_dt);\n __Pyx_XDECREF(__pyx_v_gt);\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"pycocotools/_mask.pyx\":233\n * return iou.reshape((m,n), order='F')\n * \n * def toBbox( rleObjs ): # <<<<<<<<<<<<<<\n * cdef RLEs Rs = _frString(rleObjs)\n * cdef siz n = Rs.n\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_15toBbox(PyObject *__pyx_self, PyObject *__pyx_v_rleObjs); /*proto*/\nstatic PyMethodDef __pyx_mdef_11pycocotools_5_mask_15toBbox = {\"toBbox\", (PyCFunction)__pyx_pw_11pycocotools_5_mask_15toBbox, METH_O, 0};\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_15toBbox(PyObject *__pyx_self, PyObject *__pyx_v_rleObjs) {\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"toBbox (wrapper)\", 0);\n __pyx_r = __pyx_pf_11pycocotools_5_mask_14toBbox(__pyx_self, ((PyObject *)__pyx_v_rleObjs));\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_14toBbox(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_rleObjs) {\n struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_Rs = 0;\n siz __pyx_v_n;\n BB __pyx_v__bb;\n npy_intp __pyx_v_shape[1];\n PyObject *__pyx_v_bb = NULL;\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n PyObject *__pyx_t_1 = NULL;\n PyObject *__pyx_t_2 = NULL;\n PyObject *__pyx_t_3 = NULL;\n siz __pyx_t_4;\n PyObject *__pyx_t_5 = NULL;\n PyObject *__pyx_t_6 = NULL;\n __Pyx_RefNannySetupContext(\"toBbox\", 0);\n\n /* \"pycocotools/_mask.pyx\":234\n * \n * def toBbox( rleObjs ):\n * cdef RLEs Rs = _frString(rleObjs) # <<<<<<<<<<<<<<\n * cdef siz n = Rs.n\n * cdef BB _bb = malloc(4*n* sizeof(double))\n */\n __Pyx_GetModuleGlobalName(__pyx_t_2, __pyx_n_s_frString); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 234, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __pyx_t_3 = NULL;\n if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_2))) {\n __pyx_t_3 = PyMethod_GET_SELF(__pyx_t_2);\n if (likely(__pyx_t_3)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_2);\n __Pyx_INCREF(__pyx_t_3);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_2, function);\n }\n }\n __pyx_t_1 = (__pyx_t_3) ? __Pyx_PyObject_Call2Args(__pyx_t_2, __pyx_t_3, __pyx_v_rleObjs) : __Pyx_PyObject_CallOneArg(__pyx_t_2, __pyx_v_rleObjs);\n __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;\n if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 234, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n if (!(likely(((__pyx_t_1) == Py_None) || likely(__Pyx_TypeTest(__pyx_t_1, __pyx_ptype_11pycocotools_5_mask_RLEs))))) __PYX_ERR(0, 234, __pyx_L1_error)\n __pyx_v_Rs = ((struct __pyx_obj_11pycocotools_5_mask_RLEs *)__pyx_t_1);\n __pyx_t_1 = 0;\n\n /* \"pycocotools/_mask.pyx\":235\n * def toBbox( rleObjs ):\n * cdef RLEs Rs = _frString(rleObjs)\n * cdef siz n = Rs.n # <<<<<<<<<<<<<<\n * cdef BB _bb = malloc(4*n* sizeof(double))\n * rleToBbox( Rs._R, _bb, n )\n */\n __pyx_t_1 = __Pyx_PyObject_GetAttrStr(((PyObject *)__pyx_v_Rs), __pyx_n_s_n); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 235, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_4 = __Pyx_PyInt_As_siz(__pyx_t_1); if (unlikely((__pyx_t_4 == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 235, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __pyx_v_n = __pyx_t_4;\n\n /* \"pycocotools/_mask.pyx\":236\n * cdef RLEs Rs = _frString(rleObjs)\n * cdef siz n = Rs.n\n * cdef BB _bb = malloc(4*n* sizeof(double)) # <<<<<<<<<<<<<<\n * rleToBbox( Rs._R, _bb, n )\n * cdef np.npy_intp shape[1]\n */\n __pyx_v__bb = ((BB)malloc(((4 * __pyx_v_n) * (sizeof(double)))));\n\n /* \"pycocotools/_mask.pyx\":237\n * cdef siz n = Rs.n\n * cdef BB _bb = malloc(4*n* sizeof(double))\n * rleToBbox( Rs._R, _bb, n ) # <<<<<<<<<<<<<<\n * cdef np.npy_intp shape[1]\n * shape[0] = 4*n\n */\n rleToBbox(((RLE const *)__pyx_v_Rs->_R), __pyx_v__bb, __pyx_v_n);\n\n /* \"pycocotools/_mask.pyx\":239\n * rleToBbox( Rs._R, _bb, n )\n * cdef np.npy_intp shape[1]\n * shape[0] = 4*n # <<<<<<<<<<<<<<\n * bb = np.array((1,4*n), dtype=np.double)\n * bb = np.PyArray_SimpleNewFromData(1, shape, np.NPY_DOUBLE, _bb).reshape((n, 4))\n */\n (__pyx_v_shape[0]) = (((npy_intp)4) * __pyx_v_n);\n\n /* \"pycocotools/_mask.pyx\":240\n * cdef np.npy_intp shape[1]\n * shape[0] = 4*n\n * bb = np.array((1,4*n), dtype=np.double) # <<<<<<<<<<<<<<\n * bb = np.PyArray_SimpleNewFromData(1, shape, np.NPY_DOUBLE, _bb).reshape((n, 4))\n * PyArray_ENABLEFLAGS(bb, np.NPY_OWNDATA)\n */\n __Pyx_GetModuleGlobalName(__pyx_t_1, __pyx_n_s_np); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 240, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_2 = __Pyx_PyObject_GetAttrStr(__pyx_t_1, __pyx_n_s_array); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 240, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __pyx_t_1 = __Pyx_PyInt_From_siz((4 * __pyx_v_n)); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 240, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_3 = PyTuple_New(2); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 240, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_INCREF(__pyx_int_1);\n __Pyx_GIVEREF(__pyx_int_1);\n PyTuple_SET_ITEM(__pyx_t_3, 0, __pyx_int_1);\n __Pyx_GIVEREF(__pyx_t_1);\n PyTuple_SET_ITEM(__pyx_t_3, 1, __pyx_t_1);\n __pyx_t_1 = 0;\n __pyx_t_1 = PyTuple_New(1); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 240, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_GIVEREF(__pyx_t_3);\n PyTuple_SET_ITEM(__pyx_t_1, 0, __pyx_t_3);\n __pyx_t_3 = 0;\n __pyx_t_3 = __Pyx_PyDict_NewPresized(1); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 240, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_GetModuleGlobalName(__pyx_t_5, __pyx_n_s_np); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 240, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __pyx_t_6 = __Pyx_PyObject_GetAttrStr(__pyx_t_5, __pyx_n_s_double); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 240, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_6);\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n if (PyDict_SetItem(__pyx_t_3, __pyx_n_s_dtype, __pyx_t_6) < 0) __PYX_ERR(0, 240, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;\n __pyx_t_6 = __Pyx_PyObject_Call(__pyx_t_2, __pyx_t_1, __pyx_t_3); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 240, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_6);\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __pyx_v_bb = __pyx_t_6;\n __pyx_t_6 = 0;\n\n /* \"pycocotools/_mask.pyx\":241\n * shape[0] = 4*n\n * bb = np.array((1,4*n), dtype=np.double)\n * bb = np.PyArray_SimpleNewFromData(1, shape, np.NPY_DOUBLE, _bb).reshape((n, 4)) # <<<<<<<<<<<<<<\n * PyArray_ENABLEFLAGS(bb, np.NPY_OWNDATA)\n * return bb\n */\n __pyx_t_3 = PyArray_SimpleNewFromData(1, __pyx_v_shape, NPY_DOUBLE, __pyx_v__bb); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 241, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __pyx_t_1 = __Pyx_PyObject_GetAttrStr(__pyx_t_3, __pyx_n_s_reshape); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 241, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __pyx_t_3 = __Pyx_PyInt_From_siz(__pyx_v_n); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 241, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __pyx_t_2 = PyTuple_New(2); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 241, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __Pyx_GIVEREF(__pyx_t_3);\n PyTuple_SET_ITEM(__pyx_t_2, 0, __pyx_t_3);\n __Pyx_INCREF(__pyx_int_4);\n __Pyx_GIVEREF(__pyx_int_4);\n PyTuple_SET_ITEM(__pyx_t_2, 1, __pyx_int_4);\n __pyx_t_3 = 0;\n __pyx_t_3 = NULL;\n if (CYTHON_UNPACK_METHODS && likely(PyMethod_Check(__pyx_t_1))) {\n __pyx_t_3 = PyMethod_GET_SELF(__pyx_t_1);\n if (likely(__pyx_t_3)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_1);\n __Pyx_INCREF(__pyx_t_3);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_1, function);\n }\n }\n __pyx_t_6 = (__pyx_t_3) ? __Pyx_PyObject_Call2Args(__pyx_t_1, __pyx_t_3, __pyx_t_2) : __Pyx_PyObject_CallOneArg(__pyx_t_1, __pyx_t_2);\n __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 241, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_6);\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __Pyx_DECREF_SET(__pyx_v_bb, __pyx_t_6);\n __pyx_t_6 = 0;\n\n /* \"pycocotools/_mask.pyx\":242\n * bb = np.array((1,4*n), dtype=np.double)\n * bb = np.PyArray_SimpleNewFromData(1, shape, np.NPY_DOUBLE, _bb).reshape((n, 4))\n * PyArray_ENABLEFLAGS(bb, np.NPY_OWNDATA) # <<<<<<<<<<<<<<\n * return bb\n * \n */\n if (!(likely(((__pyx_v_bb) == Py_None) || likely(__Pyx_TypeTest(__pyx_v_bb, __pyx_ptype_5numpy_ndarray))))) __PYX_ERR(0, 242, __pyx_L1_error)\n PyArray_ENABLEFLAGS(((PyArrayObject *)__pyx_v_bb), NPY_OWNDATA);\n\n /* \"pycocotools/_mask.pyx\":243\n * bb = np.PyArray_SimpleNewFromData(1, shape, np.NPY_DOUBLE, _bb).reshape((n, 4))\n * PyArray_ENABLEFLAGS(bb, np.NPY_OWNDATA)\n * return bb # <<<<<<<<<<<<<<\n * \n * def frBbox(np.ndarray[np.double_t, ndim=2] bb, siz h, siz w ):\n */\n __Pyx_XDECREF(__pyx_r);\n __Pyx_INCREF(__pyx_v_bb);\n __pyx_r = __pyx_v_bb;\n goto __pyx_L0;\n\n /* \"pycocotools/_mask.pyx\":233\n * return iou.reshape((m,n), order='F')\n * \n * def toBbox( rleObjs ): # <<<<<<<<<<<<<<\n * cdef RLEs Rs = _frString(rleObjs)\n * cdef siz n = Rs.n\n */\n\n /* function exit code */\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_1);\n __Pyx_XDECREF(__pyx_t_2);\n __Pyx_XDECREF(__pyx_t_3);\n __Pyx_XDECREF(__pyx_t_5);\n __Pyx_XDECREF(__pyx_t_6);\n __Pyx_AddTraceback(\"pycocotools._mask.toBbox\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n __pyx_L0:;\n __Pyx_XDECREF((PyObject *)__pyx_v_Rs);\n __Pyx_XDECREF(__pyx_v_bb);\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"pycocotools/_mask.pyx\":245\n * return bb\n * \n * def frBbox(np.ndarray[np.double_t, ndim=2] bb, siz h, siz w ): # <<<<<<<<<<<<<<\n * cdef siz n = bb.shape[0]\n * Rs = RLEs(n)\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_17frBbox(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds); /*proto*/\nstatic PyMethodDef __pyx_mdef_11pycocotools_5_mask_17frBbox = {\"frBbox\", (PyCFunction)(void*)(PyCFunctionWithKeywords)__pyx_pw_11pycocotools_5_mask_17frBbox, METH_VARARGS|METH_KEYWORDS, 0};\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_17frBbox(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds) {\n PyArrayObject *__pyx_v_bb = 0;\n siz __pyx_v_h;\n siz __pyx_v_w;\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"frBbox (wrapper)\", 0);\n {\n static PyObject **__pyx_pyargnames[] = {&__pyx_n_s_bb,&__pyx_n_s_h,&__pyx_n_s_w,0};\n PyObject* values[3] = {0,0,0};\n if (unlikely(__pyx_kwds)) {\n Py_ssize_t kw_args;\n const Py_ssize_t pos_args = PyTuple_GET_SIZE(__pyx_args);\n switch (pos_args) {\n case 3: values[2] = PyTuple_GET_ITEM(__pyx_args, 2);\n CYTHON_FALLTHROUGH;\n case 2: values[1] = PyTuple_GET_ITEM(__pyx_args, 1);\n CYTHON_FALLTHROUGH;\n case 1: values[0] = PyTuple_GET_ITEM(__pyx_args, 0);\n CYTHON_FALLTHROUGH;\n case 0: break;\n default: goto __pyx_L5_argtuple_error;\n }\n kw_args = PyDict_Size(__pyx_kwds);\n switch (pos_args) {\n case 0:\n if (likely((values[0] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_bb)) != 0)) kw_args--;\n else goto __pyx_L5_argtuple_error;\n CYTHON_FALLTHROUGH;\n case 1:\n if (likely((values[1] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_h)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"frBbox\", 1, 3, 3, 1); __PYX_ERR(0, 245, __pyx_L3_error)\n }\n CYTHON_FALLTHROUGH;\n case 2:\n if (likely((values[2] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_w)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"frBbox\", 1, 3, 3, 2); __PYX_ERR(0, 245, __pyx_L3_error)\n }\n }\n if (unlikely(kw_args > 0)) {\n if (unlikely(__Pyx_ParseOptionalKeywords(__pyx_kwds, __pyx_pyargnames, 0, values, pos_args, \"frBbox\") < 0)) __PYX_ERR(0, 245, __pyx_L3_error)\n }\n } else if (PyTuple_GET_SIZE(__pyx_args) != 3) {\n goto __pyx_L5_argtuple_error;\n } else {\n values[0] = PyTuple_GET_ITEM(__pyx_args, 0);\n values[1] = PyTuple_GET_ITEM(__pyx_args, 1);\n values[2] = PyTuple_GET_ITEM(__pyx_args, 2);\n }\n __pyx_v_bb = ((PyArrayObject *)values[0]);\n __pyx_v_h = __Pyx_PyInt_As_siz(values[1]); if (unlikely((__pyx_v_h == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 245, __pyx_L3_error)\n __pyx_v_w = __Pyx_PyInt_As_siz(values[2]); if (unlikely((__pyx_v_w == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 245, __pyx_L3_error)\n }\n goto __pyx_L4_argument_unpacking_done;\n __pyx_L5_argtuple_error:;\n __Pyx_RaiseArgtupleInvalid(\"frBbox\", 1, 3, 3, PyTuple_GET_SIZE(__pyx_args)); __PYX_ERR(0, 245, __pyx_L3_error)\n __pyx_L3_error:;\n __Pyx_AddTraceback(\"pycocotools._mask.frBbox\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __Pyx_RefNannyFinishContext();\n return NULL;\n __pyx_L4_argument_unpacking_done:;\n if (unlikely(!__Pyx_ArgTypeTest(((PyObject *)__pyx_v_bb), __pyx_ptype_5numpy_ndarray, 1, \"bb\", 0))) __PYX_ERR(0, 245, __pyx_L1_error)\n __pyx_r = __pyx_pf_11pycocotools_5_mask_16frBbox(__pyx_self, __pyx_v_bb, __pyx_v_h, __pyx_v_w);\n\n /* function exit code */\n goto __pyx_L0;\n __pyx_L1_error:;\n __pyx_r = NULL;\n __pyx_L0:;\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_16frBbox(CYTHON_UNUSED PyObject *__pyx_self, PyArrayObject *__pyx_v_bb, siz __pyx_v_h, siz __pyx_v_w) {\n siz __pyx_v_n;\n struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_Rs = NULL;\n PyObject *__pyx_v_objs = NULL;\n __Pyx_LocalBuf_ND __pyx_pybuffernd_bb;\n __Pyx_Buffer __pyx_pybuffer_bb;\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n PyObject *__pyx_t_1 = NULL;\n PyObject *__pyx_t_2 = NULL;\n PyObject *__pyx_t_3 = NULL;\n __Pyx_RefNannySetupContext(\"frBbox\", 0);\n __pyx_pybuffer_bb.pybuffer.buf = NULL;\n __pyx_pybuffer_bb.refcount = 0;\n __pyx_pybuffernd_bb.data = NULL;\n __pyx_pybuffernd_bb.rcbuffer = &__pyx_pybuffer_bb;\n {\n __Pyx_BufFmt_StackElem __pyx_stack[1];\n if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_bb.rcbuffer->pybuffer, (PyObject*)__pyx_v_bb, &__Pyx_TypeInfo_nn___pyx_t_5numpy_double_t, PyBUF_FORMAT| PyBUF_STRIDES, 2, 0, __pyx_stack) == -1)) __PYX_ERR(0, 245, __pyx_L1_error)\n }\n __pyx_pybuffernd_bb.diminfo[0].strides = __pyx_pybuffernd_bb.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_bb.diminfo[0].shape = __pyx_pybuffernd_bb.rcbuffer->pybuffer.shape[0]; __pyx_pybuffernd_bb.diminfo[1].strides = __pyx_pybuffernd_bb.rcbuffer->pybuffer.strides[1]; __pyx_pybuffernd_bb.diminfo[1].shape = __pyx_pybuffernd_bb.rcbuffer->pybuffer.shape[1];\n\n /* \"pycocotools/_mask.pyx\":246\n * \n * def frBbox(np.ndarray[np.double_t, ndim=2] bb, siz h, siz w ):\n * cdef siz n = bb.shape[0] # <<<<<<<<<<<<<<\n * Rs = RLEs(n)\n * rleFrBbox( Rs._R, bb.data, h, w, n )\n */\n __pyx_v_n = (__pyx_v_bb->dimensions[0]);\n\n /* \"pycocotools/_mask.pyx\":247\n * def frBbox(np.ndarray[np.double_t, ndim=2] bb, siz h, siz w ):\n * cdef siz n = bb.shape[0]\n * Rs = RLEs(n) # <<<<<<<<<<<<<<\n * rleFrBbox( Rs._R, bb.data, h, w, n )\n * objs = _toString(Rs)\n */\n __pyx_t_1 = __Pyx_PyInt_From_siz(__pyx_v_n); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 247, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_2 = __Pyx_PyObject_CallOneArg(((PyObject *)__pyx_ptype_11pycocotools_5_mask_RLEs), __pyx_t_1); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 247, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __pyx_v_Rs = ((struct __pyx_obj_11pycocotools_5_mask_RLEs *)__pyx_t_2);\n __pyx_t_2 = 0;\n\n /* \"pycocotools/_mask.pyx\":248\n * cdef siz n = bb.shape[0]\n * Rs = RLEs(n)\n * rleFrBbox( Rs._R, bb.data, h, w, n ) # <<<<<<<<<<<<<<\n * objs = _toString(Rs)\n * return objs\n */\n rleFrBbox(((RLE *)__pyx_v_Rs->_R), ((BB const )__pyx_v_bb->data), __pyx_v_h, __pyx_v_w, __pyx_v_n);\n\n /* \"pycocotools/_mask.pyx\":249\n * Rs = RLEs(n)\n * rleFrBbox( Rs._R, bb.data, h, w, n )\n * objs = _toString(Rs) # <<<<<<<<<<<<<<\n * return objs\n * \n */\n __Pyx_GetModuleGlobalName(__pyx_t_1, __pyx_n_s_toString); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 249, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_3 = NULL;\n if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_1))) {\n __pyx_t_3 = PyMethod_GET_SELF(__pyx_t_1);\n if (likely(__pyx_t_3)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_1);\n __Pyx_INCREF(__pyx_t_3);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_1, function);\n }\n }\n __pyx_t_2 = (__pyx_t_3) ? __Pyx_PyObject_Call2Args(__pyx_t_1, __pyx_t_3, ((PyObject *)__pyx_v_Rs)) : __Pyx_PyObject_CallOneArg(__pyx_t_1, ((PyObject *)__pyx_v_Rs));\n __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;\n if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 249, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __pyx_v_objs = __pyx_t_2;\n __pyx_t_2 = 0;\n\n /* \"pycocotools/_mask.pyx\":250\n * rleFrBbox( Rs._R, bb.data, h, w, n )\n * objs = _toString(Rs)\n * return objs # <<<<<<<<<<<<<<\n * \n * def frPoly( poly, siz h, siz w ):\n */\n __Pyx_XDECREF(__pyx_r);\n __Pyx_INCREF(__pyx_v_objs);\n __pyx_r = __pyx_v_objs;\n goto __pyx_L0;\n\n /* \"pycocotools/_mask.pyx\":245\n * return bb\n * \n * def frBbox(np.ndarray[np.double_t, ndim=2] bb, siz h, siz w ): # <<<<<<<<<<<<<<\n * cdef siz n = bb.shape[0]\n * Rs = RLEs(n)\n */\n\n /* function exit code */\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_1);\n __Pyx_XDECREF(__pyx_t_2);\n __Pyx_XDECREF(__pyx_t_3);\n { PyObject *__pyx_type, *__pyx_value, *__pyx_tb;\n __Pyx_PyThreadState_declare\n __Pyx_PyThreadState_assign\n __Pyx_ErrFetch(&__pyx_type, &__pyx_value, &__pyx_tb);\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_bb.rcbuffer->pybuffer);\n __Pyx_ErrRestore(__pyx_type, __pyx_value, __pyx_tb);}\n __Pyx_AddTraceback(\"pycocotools._mask.frBbox\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n goto __pyx_L2;\n __pyx_L0:;\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_bb.rcbuffer->pybuffer);\n __pyx_L2:;\n __Pyx_XDECREF((PyObject *)__pyx_v_Rs);\n __Pyx_XDECREF(__pyx_v_objs);\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"pycocotools/_mask.pyx\":252\n * return objs\n * \n * def frPoly( poly, siz h, siz w ): # <<<<<<<<<<<<<<\n * cdef np.ndarray[np.double_t, ndim=1] np_poly\n * n = len(poly)\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_19frPoly(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds); /*proto*/\nstatic PyMethodDef __pyx_mdef_11pycocotools_5_mask_19frPoly = {\"frPoly\", (PyCFunction)(void*)(PyCFunctionWithKeywords)__pyx_pw_11pycocotools_5_mask_19frPoly, METH_VARARGS|METH_KEYWORDS, 0};\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_19frPoly(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds) {\n PyObject *__pyx_v_poly = 0;\n siz __pyx_v_h;\n siz __pyx_v_w;\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"frPoly (wrapper)\", 0);\n {\n static PyObject **__pyx_pyargnames[] = {&__pyx_n_s_poly,&__pyx_n_s_h,&__pyx_n_s_w,0};\n PyObject* values[3] = {0,0,0};\n if (unlikely(__pyx_kwds)) {\n Py_ssize_t kw_args;\n const Py_ssize_t pos_args = PyTuple_GET_SIZE(__pyx_args);\n switch (pos_args) {\n case 3: values[2] = PyTuple_GET_ITEM(__pyx_args, 2);\n CYTHON_FALLTHROUGH;\n case 2: values[1] = PyTuple_GET_ITEM(__pyx_args, 1);\n CYTHON_FALLTHROUGH;\n case 1: values[0] = PyTuple_GET_ITEM(__pyx_args, 0);\n CYTHON_FALLTHROUGH;\n case 0: break;\n default: goto __pyx_L5_argtuple_error;\n }\n kw_args = PyDict_Size(__pyx_kwds);\n switch (pos_args) {\n case 0:\n if (likely((values[0] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_poly)) != 0)) kw_args--;\n else goto __pyx_L5_argtuple_error;\n CYTHON_FALLTHROUGH;\n case 1:\n if (likely((values[1] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_h)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"frPoly\", 1, 3, 3, 1); __PYX_ERR(0, 252, __pyx_L3_error)\n }\n CYTHON_FALLTHROUGH;\n case 2:\n if (likely((values[2] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_w)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"frPoly\", 1, 3, 3, 2); __PYX_ERR(0, 252, __pyx_L3_error)\n }\n }\n if (unlikely(kw_args > 0)) {\n if (unlikely(__Pyx_ParseOptionalKeywords(__pyx_kwds, __pyx_pyargnames, 0, values, pos_args, \"frPoly\") < 0)) __PYX_ERR(0, 252, __pyx_L3_error)\n }\n } else if (PyTuple_GET_SIZE(__pyx_args) != 3) {\n goto __pyx_L5_argtuple_error;\n } else {\n values[0] = PyTuple_GET_ITEM(__pyx_args, 0);\n values[1] = PyTuple_GET_ITEM(__pyx_args, 1);\n values[2] = PyTuple_GET_ITEM(__pyx_args, 2);\n }\n __pyx_v_poly = values[0];\n __pyx_v_h = __Pyx_PyInt_As_siz(values[1]); if (unlikely((__pyx_v_h == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 252, __pyx_L3_error)\n __pyx_v_w = __Pyx_PyInt_As_siz(values[2]); if (unlikely((__pyx_v_w == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 252, __pyx_L3_error)\n }\n goto __pyx_L4_argument_unpacking_done;\n __pyx_L5_argtuple_error:;\n __Pyx_RaiseArgtupleInvalid(\"frPoly\", 1, 3, 3, PyTuple_GET_SIZE(__pyx_args)); __PYX_ERR(0, 252, __pyx_L3_error)\n __pyx_L3_error:;\n __Pyx_AddTraceback(\"pycocotools._mask.frPoly\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __Pyx_RefNannyFinishContext();\n return NULL;\n __pyx_L4_argument_unpacking_done:;\n __pyx_r = __pyx_pf_11pycocotools_5_mask_18frPoly(__pyx_self, __pyx_v_poly, __pyx_v_h, __pyx_v_w);\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_18frPoly(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_poly, siz __pyx_v_h, siz __pyx_v_w) {\n PyArrayObject *__pyx_v_np_poly = 0;\n Py_ssize_t __pyx_v_n;\n struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_Rs = NULL;\n PyObject *__pyx_v_i = NULL;\n PyObject *__pyx_v_p = NULL;\n PyObject *__pyx_v_objs = NULL;\n __Pyx_LocalBuf_ND __pyx_pybuffernd_np_poly;\n __Pyx_Buffer __pyx_pybuffer_np_poly;\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n Py_ssize_t __pyx_t_1;\n PyObject *__pyx_t_2 = NULL;\n PyObject *__pyx_t_3 = NULL;\n PyObject *(*__pyx_t_4)(PyObject *);\n PyObject *__pyx_t_5 = NULL;\n PyObject *__pyx_t_6 = NULL;\n PyObject *__pyx_t_7 = NULL;\n PyObject *__pyx_t_8 = NULL;\n PyObject *__pyx_t_9 = NULL;\n PyArrayObject *__pyx_t_10 = NULL;\n int __pyx_t_11;\n PyObject *__pyx_t_12 = NULL;\n PyObject *__pyx_t_13 = NULL;\n PyObject *__pyx_t_14 = NULL;\n Py_ssize_t __pyx_t_15;\n Py_ssize_t __pyx_t_16;\n __Pyx_RefNannySetupContext(\"frPoly\", 0);\n __pyx_pybuffer_np_poly.pybuffer.buf = NULL;\n __pyx_pybuffer_np_poly.refcount = 0;\n __pyx_pybuffernd_np_poly.data = NULL;\n __pyx_pybuffernd_np_poly.rcbuffer = &__pyx_pybuffer_np_poly;\n\n /* \"pycocotools/_mask.pyx\":254\n * def frPoly( poly, siz h, siz w ):\n * cdef np.ndarray[np.double_t, ndim=1] np_poly\n * n = len(poly) # <<<<<<<<<<<<<<\n * Rs = RLEs(n)\n * for i, p in enumerate(poly):\n */\n __pyx_t_1 = PyObject_Length(__pyx_v_poly); if (unlikely(__pyx_t_1 == ((Py_ssize_t)-1))) __PYX_ERR(0, 254, __pyx_L1_error)\n __pyx_v_n = __pyx_t_1;\n\n /* \"pycocotools/_mask.pyx\":255\n * cdef np.ndarray[np.double_t, ndim=1] np_poly\n * n = len(poly)\n * Rs = RLEs(n) # <<<<<<<<<<<<<<\n * for i, p in enumerate(poly):\n * np_poly = np.array(p, dtype=np.double, order='F')\n */\n __pyx_t_2 = PyInt_FromSsize_t(__pyx_v_n); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 255, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __pyx_t_3 = __Pyx_PyObject_CallOneArg(((PyObject *)__pyx_ptype_11pycocotools_5_mask_RLEs), __pyx_t_2); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 255, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n __pyx_v_Rs = ((struct __pyx_obj_11pycocotools_5_mask_RLEs *)__pyx_t_3);\n __pyx_t_3 = 0;\n\n /* \"pycocotools/_mask.pyx\":256\n * n = len(poly)\n * Rs = RLEs(n)\n * for i, p in enumerate(poly): # <<<<<<<<<<<<<<\n * np_poly = np.array(p, dtype=np.double, order='F')\n * rleFrPoly( &Rs._R[i], np_poly.data, len(np_poly)/2, h, w )\n */\n __Pyx_INCREF(__pyx_int_0);\n __pyx_t_3 = __pyx_int_0;\n if (likely(PyList_CheckExact(__pyx_v_poly)) || PyTuple_CheckExact(__pyx_v_poly)) {\n __pyx_t_2 = __pyx_v_poly; __Pyx_INCREF(__pyx_t_2); __pyx_t_1 = 0;\n __pyx_t_4 = NULL;\n } else {\n __pyx_t_1 = -1; __pyx_t_2 = PyObject_GetIter(__pyx_v_poly); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 256, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __pyx_t_4 = Py_TYPE(__pyx_t_2)->tp_iternext; if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 256, __pyx_L1_error)\n }\n for (;;) {\n if (likely(!__pyx_t_4)) {\n if (likely(PyList_CheckExact(__pyx_t_2))) {\n if (__pyx_t_1 >= PyList_GET_SIZE(__pyx_t_2)) break;\n #if CYTHON_ASSUME_SAFE_MACROS && !CYTHON_AVOID_BORROWED_REFS\n __pyx_t_5 = PyList_GET_ITEM(__pyx_t_2, __pyx_t_1); __Pyx_INCREF(__pyx_t_5); __pyx_t_1++; if (unlikely(0 < 0)) __PYX_ERR(0, 256, __pyx_L1_error)\n #else\n __pyx_t_5 = PySequence_ITEM(__pyx_t_2, __pyx_t_1); __pyx_t_1++; if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 256, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n #endif\n } else {\n if (__pyx_t_1 >= PyTuple_GET_SIZE(__pyx_t_2)) break;\n #if CYTHON_ASSUME_SAFE_MACROS && !CYTHON_AVOID_BORROWED_REFS\n __pyx_t_5 = PyTuple_GET_ITEM(__pyx_t_2, __pyx_t_1); __Pyx_INCREF(__pyx_t_5); __pyx_t_1++; if (unlikely(0 < 0)) __PYX_ERR(0, 256, __pyx_L1_error)\n #else\n __pyx_t_5 = PySequence_ITEM(__pyx_t_2, __pyx_t_1); __pyx_t_1++; if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 256, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n #endif\n }\n } else {\n __pyx_t_5 = __pyx_t_4(__pyx_t_2);\n if (unlikely(!__pyx_t_5)) {\n PyObject* exc_type = PyErr_Occurred();\n if (exc_type) {\n if (likely(__Pyx_PyErr_GivenExceptionMatches(exc_type, PyExc_StopIteration))) PyErr_Clear();\n else __PYX_ERR(0, 256, __pyx_L1_error)\n }\n break;\n }\n __Pyx_GOTREF(__pyx_t_5);\n }\n __Pyx_XDECREF_SET(__pyx_v_p, __pyx_t_5);\n __pyx_t_5 = 0;\n __Pyx_INCREF(__pyx_t_3);\n __Pyx_XDECREF_SET(__pyx_v_i, __pyx_t_3);\n __pyx_t_5 = __Pyx_PyInt_AddObjC(__pyx_t_3, __pyx_int_1, 1, 0, 0); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 256, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_DECREF(__pyx_t_3);\n __pyx_t_3 = __pyx_t_5;\n __pyx_t_5 = 0;\n\n /* \"pycocotools/_mask.pyx\":257\n * Rs = RLEs(n)\n * for i, p in enumerate(poly):\n * np_poly = np.array(p, dtype=np.double, order='F') # <<<<<<<<<<<<<<\n * rleFrPoly( &Rs._R[i], np_poly.data, len(np_poly)/2, h, w )\n * objs = _toString(Rs)\n */\n __Pyx_GetModuleGlobalName(__pyx_t_5, __pyx_n_s_np); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 257, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __pyx_t_6 = __Pyx_PyObject_GetAttrStr(__pyx_t_5, __pyx_n_s_array); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 257, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_6);\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __pyx_t_5 = PyTuple_New(1); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 257, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_INCREF(__pyx_v_p);\n __Pyx_GIVEREF(__pyx_v_p);\n PyTuple_SET_ITEM(__pyx_t_5, 0, __pyx_v_p);\n __pyx_t_7 = __Pyx_PyDict_NewPresized(2); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 257, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_7);\n __Pyx_GetModuleGlobalName(__pyx_t_8, __pyx_n_s_np); if (unlikely(!__pyx_t_8)) __PYX_ERR(0, 257, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_8);\n __pyx_t_9 = __Pyx_PyObject_GetAttrStr(__pyx_t_8, __pyx_n_s_double); if (unlikely(!__pyx_t_9)) __PYX_ERR(0, 257, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_9);\n __Pyx_DECREF(__pyx_t_8); __pyx_t_8 = 0;\n if (PyDict_SetItem(__pyx_t_7, __pyx_n_s_dtype, __pyx_t_9) < 0) __PYX_ERR(0, 257, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_9); __pyx_t_9 = 0;\n if (PyDict_SetItem(__pyx_t_7, __pyx_n_s_order, __pyx_n_s_F) < 0) __PYX_ERR(0, 257, __pyx_L1_error)\n __pyx_t_9 = __Pyx_PyObject_Call(__pyx_t_6, __pyx_t_5, __pyx_t_7); if (unlikely(!__pyx_t_9)) __PYX_ERR(0, 257, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_9);\n __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;\n if (!(likely(((__pyx_t_9) == Py_None) || likely(__Pyx_TypeTest(__pyx_t_9, __pyx_ptype_5numpy_ndarray))))) __PYX_ERR(0, 257, __pyx_L1_error)\n __pyx_t_10 = ((PyArrayObject *)__pyx_t_9);\n {\n __Pyx_BufFmt_StackElem __pyx_stack[1];\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_np_poly.rcbuffer->pybuffer);\n __pyx_t_11 = __Pyx_GetBufferAndValidate(&__pyx_pybuffernd_np_poly.rcbuffer->pybuffer, (PyObject*)__pyx_t_10, &__Pyx_TypeInfo_nn___pyx_t_5numpy_double_t, PyBUF_FORMAT| PyBUF_STRIDES, 1, 0, __pyx_stack);\n if (unlikely(__pyx_t_11 < 0)) {\n PyErr_Fetch(&__pyx_t_12, &__pyx_t_13, &__pyx_t_14);\n if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_np_poly.rcbuffer->pybuffer, (PyObject*)__pyx_v_np_poly, &__Pyx_TypeInfo_nn___pyx_t_5numpy_double_t, PyBUF_FORMAT| PyBUF_STRIDES, 1, 0, __pyx_stack) == -1)) {\n Py_XDECREF(__pyx_t_12); Py_XDECREF(__pyx_t_13); Py_XDECREF(__pyx_t_14);\n __Pyx_RaiseBufferFallbackError();\n } else {\n PyErr_Restore(__pyx_t_12, __pyx_t_13, __pyx_t_14);\n }\n __pyx_t_12 = __pyx_t_13 = __pyx_t_14 = 0;\n }\n __pyx_pybuffernd_np_poly.diminfo[0].strides = __pyx_pybuffernd_np_poly.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_np_poly.diminfo[0].shape = __pyx_pybuffernd_np_poly.rcbuffer->pybuffer.shape[0];\n if (unlikely(__pyx_t_11 < 0)) __PYX_ERR(0, 257, __pyx_L1_error)\n }\n __pyx_t_10 = 0;\n __Pyx_XDECREF_SET(__pyx_v_np_poly, ((PyArrayObject *)__pyx_t_9));\n __pyx_t_9 = 0;\n\n /* \"pycocotools/_mask.pyx\":258\n * for i, p in enumerate(poly):\n * np_poly = np.array(p, dtype=np.double, order='F')\n * rleFrPoly( &Rs._R[i], np_poly.data, len(np_poly)/2, h, w ) # <<<<<<<<<<<<<<\n * objs = _toString(Rs)\n * return objs\n */\n __pyx_t_15 = __Pyx_PyIndex_AsSsize_t(__pyx_v_i); if (unlikely((__pyx_t_15 == (Py_ssize_t)-1) && PyErr_Occurred())) __PYX_ERR(0, 258, __pyx_L1_error)\n __pyx_t_16 = PyObject_Length(((PyObject *)__pyx_v_np_poly)); if (unlikely(__pyx_t_16 == ((Py_ssize_t)-1))) __PYX_ERR(0, 258, __pyx_L1_error)\n rleFrPoly(((RLE *)(&(__pyx_v_Rs->_R[__pyx_t_15]))), ((double const *)__pyx_v_np_poly->data), __Pyx_div_Py_ssize_t(__pyx_t_16, 2), __pyx_v_h, __pyx_v_w);\n\n /* \"pycocotools/_mask.pyx\":256\n * n = len(poly)\n * Rs = RLEs(n)\n * for i, p in enumerate(poly): # <<<<<<<<<<<<<<\n * np_poly = np.array(p, dtype=np.double, order='F')\n * rleFrPoly( &Rs._R[i], np_poly.data, len(np_poly)/2, h, w )\n */\n }\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n\n /* \"pycocotools/_mask.pyx\":259\n * np_poly = np.array(p, dtype=np.double, order='F')\n * rleFrPoly( &Rs._R[i], np_poly.data, len(np_poly)/2, h, w )\n * objs = _toString(Rs) # <<<<<<<<<<<<<<\n * return objs\n * \n */\n __Pyx_GetModuleGlobalName(__pyx_t_2, __pyx_n_s_toString); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 259, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __pyx_t_9 = NULL;\n if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_2))) {\n __pyx_t_9 = PyMethod_GET_SELF(__pyx_t_2);\n if (likely(__pyx_t_9)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_2);\n __Pyx_INCREF(__pyx_t_9);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_2, function);\n }\n }\n __pyx_t_3 = (__pyx_t_9) ? __Pyx_PyObject_Call2Args(__pyx_t_2, __pyx_t_9, ((PyObject *)__pyx_v_Rs)) : __Pyx_PyObject_CallOneArg(__pyx_t_2, ((PyObject *)__pyx_v_Rs));\n __Pyx_XDECREF(__pyx_t_9); __pyx_t_9 = 0;\n if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 259, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n __pyx_v_objs = __pyx_t_3;\n __pyx_t_3 = 0;\n\n /* \"pycocotools/_mask.pyx\":260\n * rleFrPoly( &Rs._R[i], np_poly.data, len(np_poly)/2, h, w )\n * objs = _toString(Rs)\n * return objs # <<<<<<<<<<<<<<\n * \n * def frUncompressedRLE(ucRles, siz h, siz w):\n */\n __Pyx_XDECREF(__pyx_r);\n __Pyx_INCREF(__pyx_v_objs);\n __pyx_r = __pyx_v_objs;\n goto __pyx_L0;\n\n /* \"pycocotools/_mask.pyx\":252\n * return objs\n * \n * def frPoly( poly, siz h, siz w ): # <<<<<<<<<<<<<<\n * cdef np.ndarray[np.double_t, ndim=1] np_poly\n * n = len(poly)\n */\n\n /* function exit code */\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_2);\n __Pyx_XDECREF(__pyx_t_3);\n __Pyx_XDECREF(__pyx_t_5);\n __Pyx_XDECREF(__pyx_t_6);\n __Pyx_XDECREF(__pyx_t_7);\n __Pyx_XDECREF(__pyx_t_8);\n __Pyx_XDECREF(__pyx_t_9);\n { PyObject *__pyx_type, *__pyx_value, *__pyx_tb;\n __Pyx_PyThreadState_declare\n __Pyx_PyThreadState_assign\n __Pyx_ErrFetch(&__pyx_type, &__pyx_value, &__pyx_tb);\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_np_poly.rcbuffer->pybuffer);\n __Pyx_ErrRestore(__pyx_type, __pyx_value, __pyx_tb);}\n __Pyx_AddTraceback(\"pycocotools._mask.frPoly\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n goto __pyx_L2;\n __pyx_L0:;\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_np_poly.rcbuffer->pybuffer);\n __pyx_L2:;\n __Pyx_XDECREF((PyObject *)__pyx_v_np_poly);\n __Pyx_XDECREF((PyObject *)__pyx_v_Rs);\n __Pyx_XDECREF(__pyx_v_i);\n __Pyx_XDECREF(__pyx_v_p);\n __Pyx_XDECREF(__pyx_v_objs);\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"pycocotools/_mask.pyx\":262\n * return objs\n * \n * def frUncompressedRLE(ucRles, siz h, siz w): # <<<<<<<<<<<<<<\n * cdef np.ndarray[np.uint32_t, ndim=1] cnts\n * cdef RLE R\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_21frUncompressedRLE(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds); /*proto*/\nstatic PyMethodDef __pyx_mdef_11pycocotools_5_mask_21frUncompressedRLE = {\"frUncompressedRLE\", (PyCFunction)(void*)(PyCFunctionWithKeywords)__pyx_pw_11pycocotools_5_mask_21frUncompressedRLE, METH_VARARGS|METH_KEYWORDS, 0};\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_21frUncompressedRLE(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds) {\n PyObject *__pyx_v_ucRles = 0;\n CYTHON_UNUSED siz __pyx_v_h;\n CYTHON_UNUSED siz __pyx_v_w;\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"frUncompressedRLE (wrapper)\", 0);\n {\n static PyObject **__pyx_pyargnames[] = {&__pyx_n_s_ucRles,&__pyx_n_s_h,&__pyx_n_s_w,0};\n PyObject* values[3] = {0,0,0};\n if (unlikely(__pyx_kwds)) {\n Py_ssize_t kw_args;\n const Py_ssize_t pos_args = PyTuple_GET_SIZE(__pyx_args);\n switch (pos_args) {\n case 3: values[2] = PyTuple_GET_ITEM(__pyx_args, 2);\n CYTHON_FALLTHROUGH;\n case 2: values[1] = PyTuple_GET_ITEM(__pyx_args, 1);\n CYTHON_FALLTHROUGH;\n case 1: values[0] = PyTuple_GET_ITEM(__pyx_args, 0);\n CYTHON_FALLTHROUGH;\n case 0: break;\n default: goto __pyx_L5_argtuple_error;\n }\n kw_args = PyDict_Size(__pyx_kwds);\n switch (pos_args) {\n case 0:\n if (likely((values[0] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_ucRles)) != 0)) kw_args--;\n else goto __pyx_L5_argtuple_error;\n CYTHON_FALLTHROUGH;\n case 1:\n if (likely((values[1] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_h)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"frUncompressedRLE\", 1, 3, 3, 1); __PYX_ERR(0, 262, __pyx_L3_error)\n }\n CYTHON_FALLTHROUGH;\n case 2:\n if (likely((values[2] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_w)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"frUncompressedRLE\", 1, 3, 3, 2); __PYX_ERR(0, 262, __pyx_L3_error)\n }\n }\n if (unlikely(kw_args > 0)) {\n if (unlikely(__Pyx_ParseOptionalKeywords(__pyx_kwds, __pyx_pyargnames, 0, values, pos_args, \"frUncompressedRLE\") < 0)) __PYX_ERR(0, 262, __pyx_L3_error)\n }\n } else if (PyTuple_GET_SIZE(__pyx_args) != 3) {\n goto __pyx_L5_argtuple_error;\n } else {\n values[0] = PyTuple_GET_ITEM(__pyx_args, 0);\n values[1] = PyTuple_GET_ITEM(__pyx_args, 1);\n values[2] = PyTuple_GET_ITEM(__pyx_args, 2);\n }\n __pyx_v_ucRles = values[0];\n __pyx_v_h = __Pyx_PyInt_As_siz(values[1]); if (unlikely((__pyx_v_h == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 262, __pyx_L3_error)\n __pyx_v_w = __Pyx_PyInt_As_siz(values[2]); if (unlikely((__pyx_v_w == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 262, __pyx_L3_error)\n }\n goto __pyx_L4_argument_unpacking_done;\n __pyx_L5_argtuple_error:;\n __Pyx_RaiseArgtupleInvalid(\"frUncompressedRLE\", 1, 3, 3, PyTuple_GET_SIZE(__pyx_args)); __PYX_ERR(0, 262, __pyx_L3_error)\n __pyx_L3_error:;\n __Pyx_AddTraceback(\"pycocotools._mask.frUncompressedRLE\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __Pyx_RefNannyFinishContext();\n return NULL;\n __pyx_L4_argument_unpacking_done:;\n __pyx_r = __pyx_pf_11pycocotools_5_mask_20frUncompressedRLE(__pyx_self, __pyx_v_ucRles, __pyx_v_h, __pyx_v_w);\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_20frUncompressedRLE(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_ucRles, CYTHON_UNUSED siz __pyx_v_h, CYTHON_UNUSED siz __pyx_v_w) {\n PyArrayObject *__pyx_v_cnts = 0;\n RLE __pyx_v_R;\n uint *__pyx_v_data;\n Py_ssize_t __pyx_v_n;\n PyObject *__pyx_v_objs = NULL;\n Py_ssize_t __pyx_v_i;\n struct __pyx_obj_11pycocotools_5_mask_RLEs *__pyx_v_Rs = NULL;\n Py_ssize_t __pyx_v_j;\n __Pyx_LocalBuf_ND __pyx_pybuffernd_cnts;\n __Pyx_Buffer __pyx_pybuffer_cnts;\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n Py_ssize_t __pyx_t_1;\n PyObject *__pyx_t_2 = NULL;\n Py_ssize_t __pyx_t_3;\n Py_ssize_t __pyx_t_4;\n PyObject *__pyx_t_5 = NULL;\n PyObject *__pyx_t_6 = NULL;\n PyObject *__pyx_t_7 = NULL;\n PyObject *__pyx_t_8 = NULL;\n PyArrayObject *__pyx_t_9 = NULL;\n int __pyx_t_10;\n PyObject *__pyx_t_11 = NULL;\n PyObject *__pyx_t_12 = NULL;\n PyObject *__pyx_t_13 = NULL;\n Py_ssize_t __pyx_t_14;\n Py_ssize_t __pyx_t_15;\n Py_ssize_t __pyx_t_16;\n Py_ssize_t __pyx_t_17;\n RLE __pyx_t_18;\n siz __pyx_t_19;\n int __pyx_t_20;\n __Pyx_RefNannySetupContext(\"frUncompressedRLE\", 0);\n __pyx_pybuffer_cnts.pybuffer.buf = NULL;\n __pyx_pybuffer_cnts.refcount = 0;\n __pyx_pybuffernd_cnts.data = NULL;\n __pyx_pybuffernd_cnts.rcbuffer = &__pyx_pybuffer_cnts;\n\n /* \"pycocotools/_mask.pyx\":266\n * cdef RLE R\n * cdef uint *data\n * n = len(ucRles) # <<<<<<<<<<<<<<\n * objs = []\n * for i in range(n):\n */\n __pyx_t_1 = PyObject_Length(__pyx_v_ucRles); if (unlikely(__pyx_t_1 == ((Py_ssize_t)-1))) __PYX_ERR(0, 266, __pyx_L1_error)\n __pyx_v_n = __pyx_t_1;\n\n /* \"pycocotools/_mask.pyx\":267\n * cdef uint *data\n * n = len(ucRles)\n * objs = [] # <<<<<<<<<<<<<<\n * for i in range(n):\n * Rs = RLEs(1)\n */\n __pyx_t_2 = PyList_New(0); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 267, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __pyx_v_objs = ((PyObject*)__pyx_t_2);\n __pyx_t_2 = 0;\n\n /* \"pycocotools/_mask.pyx\":268\n * n = len(ucRles)\n * objs = []\n * for i in range(n): # <<<<<<<<<<<<<<\n * Rs = RLEs(1)\n * cnts = np.array(ucRles[i]['counts'], dtype=np.uint32)\n */\n __pyx_t_1 = __pyx_v_n;\n __pyx_t_3 = __pyx_t_1;\n for (__pyx_t_4 = 0; __pyx_t_4 < __pyx_t_3; __pyx_t_4+=1) {\n __pyx_v_i = __pyx_t_4;\n\n /* \"pycocotools/_mask.pyx\":269\n * objs = []\n * for i in range(n):\n * Rs = RLEs(1) # <<<<<<<<<<<<<<\n * cnts = np.array(ucRles[i]['counts'], dtype=np.uint32)\n * # time for malloc can be saved here but it's fine\n */\n __pyx_t_2 = __Pyx_PyObject_Call(((PyObject *)__pyx_ptype_11pycocotools_5_mask_RLEs), __pyx_tuple__5, NULL); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 269, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __Pyx_XDECREF_SET(__pyx_v_Rs, ((struct __pyx_obj_11pycocotools_5_mask_RLEs *)__pyx_t_2));\n __pyx_t_2 = 0;\n\n /* \"pycocotools/_mask.pyx\":270\n * for i in range(n):\n * Rs = RLEs(1)\n * cnts = np.array(ucRles[i]['counts'], dtype=np.uint32) # <<<<<<<<<<<<<<\n * # time for malloc can be saved here but it's fine\n * data = malloc(len(cnts)* sizeof(uint))\n */\n __Pyx_GetModuleGlobalName(__pyx_t_2, __pyx_n_s_np); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 270, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __pyx_t_5 = __Pyx_PyObject_GetAttrStr(__pyx_t_2, __pyx_n_s_array); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 270, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n __pyx_t_2 = __Pyx_GetItemInt(__pyx_v_ucRles, __pyx_v_i, Py_ssize_t, 1, PyInt_FromSsize_t, 0, 1, 1); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 270, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __pyx_t_6 = __Pyx_PyObject_Dict_GetItem(__pyx_t_2, __pyx_n_s_counts); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 270, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_6);\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n __pyx_t_2 = PyTuple_New(1); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 270, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_2);\n __Pyx_GIVEREF(__pyx_t_6);\n PyTuple_SET_ITEM(__pyx_t_2, 0, __pyx_t_6);\n __pyx_t_6 = 0;\n __pyx_t_6 = __Pyx_PyDict_NewPresized(1); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 270, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_6);\n __Pyx_GetModuleGlobalName(__pyx_t_7, __pyx_n_s_np); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 270, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_7);\n __pyx_t_8 = __Pyx_PyObject_GetAttrStr(__pyx_t_7, __pyx_n_s_uint32); if (unlikely(!__pyx_t_8)) __PYX_ERR(0, 270, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_8);\n __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;\n if (PyDict_SetItem(__pyx_t_6, __pyx_n_s_dtype, __pyx_t_8) < 0) __PYX_ERR(0, 270, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_8); __pyx_t_8 = 0;\n __pyx_t_8 = __Pyx_PyObject_Call(__pyx_t_5, __pyx_t_2, __pyx_t_6); if (unlikely(!__pyx_t_8)) __PYX_ERR(0, 270, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_8);\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;\n __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;\n if (!(likely(((__pyx_t_8) == Py_None) || likely(__Pyx_TypeTest(__pyx_t_8, __pyx_ptype_5numpy_ndarray))))) __PYX_ERR(0, 270, __pyx_L1_error)\n __pyx_t_9 = ((PyArrayObject *)__pyx_t_8);\n {\n __Pyx_BufFmt_StackElem __pyx_stack[1];\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_cnts.rcbuffer->pybuffer);\n __pyx_t_10 = __Pyx_GetBufferAndValidate(&__pyx_pybuffernd_cnts.rcbuffer->pybuffer, (PyObject*)__pyx_t_9, &__Pyx_TypeInfo_nn___pyx_t_5numpy_uint32_t, PyBUF_FORMAT| PyBUF_STRIDES, 1, 0, __pyx_stack);\n if (unlikely(__pyx_t_10 < 0)) {\n PyErr_Fetch(&__pyx_t_11, &__pyx_t_12, &__pyx_t_13);\n if (unlikely(__Pyx_GetBufferAndValidate(&__pyx_pybuffernd_cnts.rcbuffer->pybuffer, (PyObject*)__pyx_v_cnts, &__Pyx_TypeInfo_nn___pyx_t_5numpy_uint32_t, PyBUF_FORMAT| PyBUF_STRIDES, 1, 0, __pyx_stack) == -1)) {\n Py_XDECREF(__pyx_t_11); Py_XDECREF(__pyx_t_12); Py_XDECREF(__pyx_t_13);\n __Pyx_RaiseBufferFallbackError();\n } else {\n PyErr_Restore(__pyx_t_11, __pyx_t_12, __pyx_t_13);\n }\n __pyx_t_11 = __pyx_t_12 = __pyx_t_13 = 0;\n }\n __pyx_pybuffernd_cnts.diminfo[0].strides = __pyx_pybuffernd_cnts.rcbuffer->pybuffer.strides[0]; __pyx_pybuffernd_cnts.diminfo[0].shape = __pyx_pybuffernd_cnts.rcbuffer->pybuffer.shape[0];\n if (unlikely(__pyx_t_10 < 0)) __PYX_ERR(0, 270, __pyx_L1_error)\n }\n __pyx_t_9 = 0;\n __Pyx_XDECREF_SET(__pyx_v_cnts, ((PyArrayObject *)__pyx_t_8));\n __pyx_t_8 = 0;\n\n /* \"pycocotools/_mask.pyx\":272\n * cnts = np.array(ucRles[i]['counts'], dtype=np.uint32)\n * # time for malloc can be saved here but it's fine\n * data = malloc(len(cnts)* sizeof(uint)) # <<<<<<<<<<<<<<\n * for j in range(len(cnts)):\n * data[j] = cnts[j]\n */\n __pyx_t_14 = PyObject_Length(((PyObject *)__pyx_v_cnts)); if (unlikely(__pyx_t_14 == ((Py_ssize_t)-1))) __PYX_ERR(0, 272, __pyx_L1_error)\n __pyx_v_data = ((uint *)malloc((__pyx_t_14 * (sizeof(unsigned int)))));\n\n /* \"pycocotools/_mask.pyx\":273\n * # time for malloc can be saved here but it's fine\n * data = malloc(len(cnts)* sizeof(uint))\n * for j in range(len(cnts)): # <<<<<<<<<<<<<<\n * data[j] = cnts[j]\n * R = RLE(ucRles[i]['size'][0], ucRles[i]['size'][1], len(cnts), data)\n */\n __pyx_t_14 = PyObject_Length(((PyObject *)__pyx_v_cnts)); if (unlikely(__pyx_t_14 == ((Py_ssize_t)-1))) __PYX_ERR(0, 273, __pyx_L1_error)\n __pyx_t_15 = __pyx_t_14;\n for (__pyx_t_16 = 0; __pyx_t_16 < __pyx_t_15; __pyx_t_16+=1) {\n __pyx_v_j = __pyx_t_16;\n\n /* \"pycocotools/_mask.pyx\":274\n * data = malloc(len(cnts)* sizeof(uint))\n * for j in range(len(cnts)):\n * data[j] = cnts[j] # <<<<<<<<<<<<<<\n * R = RLE(ucRles[i]['size'][0], ucRles[i]['size'][1], len(cnts), data)\n * Rs._R[0] = R\n */\n __pyx_t_17 = __pyx_v_j;\n __pyx_t_10 = -1;\n if (__pyx_t_17 < 0) {\n __pyx_t_17 += __pyx_pybuffernd_cnts.diminfo[0].shape;\n if (unlikely(__pyx_t_17 < 0)) __pyx_t_10 = 0;\n } else if (unlikely(__pyx_t_17 >= __pyx_pybuffernd_cnts.diminfo[0].shape)) __pyx_t_10 = 0;\n if (unlikely(__pyx_t_10 != -1)) {\n __Pyx_RaiseBufferIndexError(__pyx_t_10);\n __PYX_ERR(0, 274, __pyx_L1_error)\n }\n (__pyx_v_data[__pyx_v_j]) = ((uint)(*__Pyx_BufPtrStrided1d(__pyx_t_5numpy_uint32_t *, __pyx_pybuffernd_cnts.rcbuffer->pybuffer.buf, __pyx_t_17, __pyx_pybuffernd_cnts.diminfo[0].strides)));\n }\n\n /* \"pycocotools/_mask.pyx\":275\n * for j in range(len(cnts)):\n * data[j] = cnts[j]\n * R = RLE(ucRles[i]['size'][0], ucRles[i]['size'][1], len(cnts), data) # <<<<<<<<<<<<<<\n * Rs._R[0] = R\n * objs.append(_toString(Rs)[0])\n */\n __pyx_t_8 = __Pyx_GetItemInt(__pyx_v_ucRles, __pyx_v_i, Py_ssize_t, 1, PyInt_FromSsize_t, 0, 1, 1); if (unlikely(!__pyx_t_8)) __PYX_ERR(0, 275, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_8);\n __pyx_t_6 = __Pyx_PyObject_Dict_GetItem(__pyx_t_8, __pyx_n_s_size); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 275, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_6);\n __Pyx_DECREF(__pyx_t_8); __pyx_t_8 = 0;\n __pyx_t_8 = __Pyx_GetItemInt(__pyx_t_6, 0, long, 1, __Pyx_PyInt_From_long, 0, 0, 1); if (unlikely(!__pyx_t_8)) __PYX_ERR(0, 275, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_8);\n __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;\n __pyx_t_19 = __Pyx_PyInt_As_siz(__pyx_t_8); if (unlikely((__pyx_t_19 == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 275, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_8); __pyx_t_8 = 0;\n __pyx_t_18.h = __pyx_t_19;\n __pyx_t_8 = __Pyx_GetItemInt(__pyx_v_ucRles, __pyx_v_i, Py_ssize_t, 1, PyInt_FromSsize_t, 0, 1, 1); if (unlikely(!__pyx_t_8)) __PYX_ERR(0, 275, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_8);\n __pyx_t_6 = __Pyx_PyObject_Dict_GetItem(__pyx_t_8, __pyx_n_s_size); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 275, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_6);\n __Pyx_DECREF(__pyx_t_8); __pyx_t_8 = 0;\n __pyx_t_8 = __Pyx_GetItemInt(__pyx_t_6, 1, long, 1, __Pyx_PyInt_From_long, 0, 0, 1); if (unlikely(!__pyx_t_8)) __PYX_ERR(0, 275, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_8);\n __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;\n __pyx_t_19 = __Pyx_PyInt_As_siz(__pyx_t_8); if (unlikely((__pyx_t_19 == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 275, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_8); __pyx_t_8 = 0;\n __pyx_t_18.w = __pyx_t_19;\n __pyx_t_14 = PyObject_Length(((PyObject *)__pyx_v_cnts)); if (unlikely(__pyx_t_14 == ((Py_ssize_t)-1))) __PYX_ERR(0, 275, __pyx_L1_error)\n __pyx_t_18.m = __pyx_t_14;\n __pyx_t_18.cnts = ((uint *)__pyx_v_data);\n __pyx_v_R = __pyx_t_18;\n\n /* \"pycocotools/_mask.pyx\":276\n * data[j] = cnts[j]\n * R = RLE(ucRles[i]['size'][0], ucRles[i]['size'][1], len(cnts), data)\n * Rs._R[0] = R # <<<<<<<<<<<<<<\n * objs.append(_toString(Rs)[0])\n * return objs\n */\n (__pyx_v_Rs->_R[0]) = __pyx_v_R;\n\n /* \"pycocotools/_mask.pyx\":277\n * R = RLE(ucRles[i]['size'][0], ucRles[i]['size'][1], len(cnts), data)\n * Rs._R[0] = R\n * objs.append(_toString(Rs)[0]) # <<<<<<<<<<<<<<\n * return objs\n * \n */\n __Pyx_GetModuleGlobalName(__pyx_t_6, __pyx_n_s_toString); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 277, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_6);\n __pyx_t_2 = NULL;\n if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_6))) {\n __pyx_t_2 = PyMethod_GET_SELF(__pyx_t_6);\n if (likely(__pyx_t_2)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_6);\n __Pyx_INCREF(__pyx_t_2);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_6, function);\n }\n }\n __pyx_t_8 = (__pyx_t_2) ? __Pyx_PyObject_Call2Args(__pyx_t_6, __pyx_t_2, ((PyObject *)__pyx_v_Rs)) : __Pyx_PyObject_CallOneArg(__pyx_t_6, ((PyObject *)__pyx_v_Rs));\n __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;\n if (unlikely(!__pyx_t_8)) __PYX_ERR(0, 277, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_8);\n __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;\n __pyx_t_6 = __Pyx_GetItemInt(__pyx_t_8, 0, long, 1, __Pyx_PyInt_From_long, 0, 0, 1); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 277, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_6);\n __Pyx_DECREF(__pyx_t_8); __pyx_t_8 = 0;\n __pyx_t_20 = __Pyx_PyList_Append(__pyx_v_objs, __pyx_t_6); if (unlikely(__pyx_t_20 == ((int)-1))) __PYX_ERR(0, 277, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;\n }\n\n /* \"pycocotools/_mask.pyx\":278\n * Rs._R[0] = R\n * objs.append(_toString(Rs)[0])\n * return objs # <<<<<<<<<<<<<<\n * \n * def frPyObjects(pyobj, siz h, w):\n */\n __Pyx_XDECREF(__pyx_r);\n __Pyx_INCREF(__pyx_v_objs);\n __pyx_r = __pyx_v_objs;\n goto __pyx_L0;\n\n /* \"pycocotools/_mask.pyx\":262\n * return objs\n * \n * def frUncompressedRLE(ucRles, siz h, siz w): # <<<<<<<<<<<<<<\n * cdef np.ndarray[np.uint32_t, ndim=1] cnts\n * cdef RLE R\n */\n\n /* function exit code */\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_2);\n __Pyx_XDECREF(__pyx_t_5);\n __Pyx_XDECREF(__pyx_t_6);\n __Pyx_XDECREF(__pyx_t_7);\n __Pyx_XDECREF(__pyx_t_8);\n { PyObject *__pyx_type, *__pyx_value, *__pyx_tb;\n __Pyx_PyThreadState_declare\n __Pyx_PyThreadState_assign\n __Pyx_ErrFetch(&__pyx_type, &__pyx_value, &__pyx_tb);\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_cnts.rcbuffer->pybuffer);\n __Pyx_ErrRestore(__pyx_type, __pyx_value, __pyx_tb);}\n __Pyx_AddTraceback(\"pycocotools._mask.frUncompressedRLE\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n goto __pyx_L2;\n __pyx_L0:;\n __Pyx_SafeReleaseBuffer(&__pyx_pybuffernd_cnts.rcbuffer->pybuffer);\n __pyx_L2:;\n __Pyx_XDECREF((PyObject *)__pyx_v_cnts);\n __Pyx_XDECREF(__pyx_v_objs);\n __Pyx_XDECREF((PyObject *)__pyx_v_Rs);\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"pycocotools/_mask.pyx\":280\n * return objs\n * \n * def frPyObjects(pyobj, siz h, w): # <<<<<<<<<<<<<<\n * if type(pyobj) == np.ndarray:\n * objs = frBbox(pyobj, h, w )\n */\n\n/* Python wrapper */\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_23frPyObjects(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds); /*proto*/\nstatic PyMethodDef __pyx_mdef_11pycocotools_5_mask_23frPyObjects = {\"frPyObjects\", (PyCFunction)(void*)(PyCFunctionWithKeywords)__pyx_pw_11pycocotools_5_mask_23frPyObjects, METH_VARARGS|METH_KEYWORDS, 0};\nstatic PyObject *__pyx_pw_11pycocotools_5_mask_23frPyObjects(PyObject *__pyx_self, PyObject *__pyx_args, PyObject *__pyx_kwds) {\n PyObject *__pyx_v_pyobj = 0;\n siz __pyx_v_h;\n PyObject *__pyx_v_w = 0;\n PyObject *__pyx_r = 0;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"frPyObjects (wrapper)\", 0);\n {\n static PyObject **__pyx_pyargnames[] = {&__pyx_n_s_pyobj,&__pyx_n_s_h,&__pyx_n_s_w,0};\n PyObject* values[3] = {0,0,0};\n if (unlikely(__pyx_kwds)) {\n Py_ssize_t kw_args;\n const Py_ssize_t pos_args = PyTuple_GET_SIZE(__pyx_args);\n switch (pos_args) {\n case 3: values[2] = PyTuple_GET_ITEM(__pyx_args, 2);\n CYTHON_FALLTHROUGH;\n case 2: values[1] = PyTuple_GET_ITEM(__pyx_args, 1);\n CYTHON_FALLTHROUGH;\n case 1: values[0] = PyTuple_GET_ITEM(__pyx_args, 0);\n CYTHON_FALLTHROUGH;\n case 0: break;\n default: goto __pyx_L5_argtuple_error;\n }\n kw_args = PyDict_Size(__pyx_kwds);\n switch (pos_args) {\n case 0:\n if (likely((values[0] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_pyobj)) != 0)) kw_args--;\n else goto __pyx_L5_argtuple_error;\n CYTHON_FALLTHROUGH;\n case 1:\n if (likely((values[1] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_h)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"frPyObjects\", 1, 3, 3, 1); __PYX_ERR(0, 280, __pyx_L3_error)\n }\n CYTHON_FALLTHROUGH;\n case 2:\n if (likely((values[2] = __Pyx_PyDict_GetItemStr(__pyx_kwds, __pyx_n_s_w)) != 0)) kw_args--;\n else {\n __Pyx_RaiseArgtupleInvalid(\"frPyObjects\", 1, 3, 3, 2); __PYX_ERR(0, 280, __pyx_L3_error)\n }\n }\n if (unlikely(kw_args > 0)) {\n if (unlikely(__Pyx_ParseOptionalKeywords(__pyx_kwds, __pyx_pyargnames, 0, values, pos_args, \"frPyObjects\") < 0)) __PYX_ERR(0, 280, __pyx_L3_error)\n }\n } else if (PyTuple_GET_SIZE(__pyx_args) != 3) {\n goto __pyx_L5_argtuple_error;\n } else {\n values[0] = PyTuple_GET_ITEM(__pyx_args, 0);\n values[1] = PyTuple_GET_ITEM(__pyx_args, 1);\n values[2] = PyTuple_GET_ITEM(__pyx_args, 2);\n }\n __pyx_v_pyobj = values[0];\n __pyx_v_h = __Pyx_PyInt_As_siz(values[1]); if (unlikely((__pyx_v_h == ((siz)-1)) && PyErr_Occurred())) __PYX_ERR(0, 280, __pyx_L3_error)\n __pyx_v_w = values[2];\n }\n goto __pyx_L4_argument_unpacking_done;\n __pyx_L5_argtuple_error:;\n __Pyx_RaiseArgtupleInvalid(\"frPyObjects\", 1, 3, 3, PyTuple_GET_SIZE(__pyx_args)); __PYX_ERR(0, 280, __pyx_L3_error)\n __pyx_L3_error:;\n __Pyx_AddTraceback(\"pycocotools._mask.frPyObjects\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __Pyx_RefNannyFinishContext();\n return NULL;\n __pyx_L4_argument_unpacking_done:;\n __pyx_r = __pyx_pf_11pycocotools_5_mask_22frPyObjects(__pyx_self, __pyx_v_pyobj, __pyx_v_h, __pyx_v_w);\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic PyObject *__pyx_pf_11pycocotools_5_mask_22frPyObjects(CYTHON_UNUSED PyObject *__pyx_self, PyObject *__pyx_v_pyobj, siz __pyx_v_h, PyObject *__pyx_v_w) {\n PyObject *__pyx_v_objs = NULL;\n PyObject *__pyx_r = NULL;\n __Pyx_RefNannyDeclarations\n PyObject *__pyx_t_1 = NULL;\n int __pyx_t_2;\n PyObject *__pyx_t_3 = NULL;\n PyObject *__pyx_t_4 = NULL;\n PyObject *__pyx_t_5 = NULL;\n int __pyx_t_6;\n PyObject *__pyx_t_7 = NULL;\n int __pyx_t_8;\n Py_ssize_t __pyx_t_9;\n __Pyx_RefNannySetupContext(\"frPyObjects\", 0);\n\n /* \"pycocotools/_mask.pyx\":281\n * \n * def frPyObjects(pyobj, siz h, w):\n * if type(pyobj) == np.ndarray: # <<<<<<<<<<<<<<\n * objs = frBbox(pyobj, h, w )\n * elif type(pyobj) == list and len(pyobj[0]) == 4:\n */\n __pyx_t_1 = PyObject_RichCompare(((PyObject *)Py_TYPE(__pyx_v_pyobj)), ((PyObject *)__pyx_ptype_5numpy_ndarray), Py_EQ); __Pyx_XGOTREF(__pyx_t_1); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 281, __pyx_L1_error)\n __pyx_t_2 = __Pyx_PyObject_IsTrue(__pyx_t_1); if (unlikely(__pyx_t_2 < 0)) __PYX_ERR(0, 281, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n if (__pyx_t_2) {\n\n /* \"pycocotools/_mask.pyx\":282\n * def frPyObjects(pyobj, siz h, w):\n * if type(pyobj) == np.ndarray:\n * objs = frBbox(pyobj, h, w ) # <<<<<<<<<<<<<<\n * elif type(pyobj) == list and len(pyobj[0]) == 4:\n * objs = frBbox(pyobj, h, w )\n */\n __Pyx_GetModuleGlobalName(__pyx_t_3, __pyx_n_s_frBbox); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 282, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __pyx_t_4 = __Pyx_PyInt_From_siz(__pyx_v_h); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 282, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __pyx_t_5 = NULL;\n __pyx_t_6 = 0;\n if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_3))) {\n __pyx_t_5 = PyMethod_GET_SELF(__pyx_t_3);\n if (likely(__pyx_t_5)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_3);\n __Pyx_INCREF(__pyx_t_5);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_3, function);\n __pyx_t_6 = 1;\n }\n }\n #if CYTHON_FAST_PYCALL\n if (PyFunction_Check(__pyx_t_3)) {\n PyObject *__pyx_temp[4] = {__pyx_t_5, __pyx_v_pyobj, __pyx_t_4, __pyx_v_w};\n __pyx_t_1 = __Pyx_PyFunction_FastCall(__pyx_t_3, __pyx_temp+1-__pyx_t_6, 3+__pyx_t_6); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 282, __pyx_L1_error)\n __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;\n } else\n #endif\n #if CYTHON_FAST_PYCCALL\n if (__Pyx_PyFastCFunction_Check(__pyx_t_3)) {\n PyObject *__pyx_temp[4] = {__pyx_t_5, __pyx_v_pyobj, __pyx_t_4, __pyx_v_w};\n __pyx_t_1 = __Pyx_PyCFunction_FastCall(__pyx_t_3, __pyx_temp+1-__pyx_t_6, 3+__pyx_t_6); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 282, __pyx_L1_error)\n __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;\n } else\n #endif\n {\n __pyx_t_7 = PyTuple_New(3+__pyx_t_6); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 282, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_7);\n if (__pyx_t_5) {\n __Pyx_GIVEREF(__pyx_t_5); PyTuple_SET_ITEM(__pyx_t_7, 0, __pyx_t_5); __pyx_t_5 = NULL;\n }\n __Pyx_INCREF(__pyx_v_pyobj);\n __Pyx_GIVEREF(__pyx_v_pyobj);\n PyTuple_SET_ITEM(__pyx_t_7, 0+__pyx_t_6, __pyx_v_pyobj);\n __Pyx_GIVEREF(__pyx_t_4);\n PyTuple_SET_ITEM(__pyx_t_7, 1+__pyx_t_6, __pyx_t_4);\n __Pyx_INCREF(__pyx_v_w);\n __Pyx_GIVEREF(__pyx_v_w);\n PyTuple_SET_ITEM(__pyx_t_7, 2+__pyx_t_6, __pyx_v_w);\n __pyx_t_4 = 0;\n __pyx_t_1 = __Pyx_PyObject_Call(__pyx_t_3, __pyx_t_7, NULL); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 282, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;\n }\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __pyx_v_objs = __pyx_t_1;\n __pyx_t_1 = 0;\n\n /* \"pycocotools/_mask.pyx\":281\n * \n * def frPyObjects(pyobj, siz h, w):\n * if type(pyobj) == np.ndarray: # <<<<<<<<<<<<<<\n * objs = frBbox(pyobj, h, w )\n * elif type(pyobj) == list and len(pyobj[0]) == 4:\n */\n goto __pyx_L3;\n }\n\n /* \"pycocotools/_mask.pyx\":283\n * if type(pyobj) == np.ndarray:\n * objs = frBbox(pyobj, h, w )\n * elif type(pyobj) == list and len(pyobj[0]) == 4: # <<<<<<<<<<<<<<\n * objs = frBbox(pyobj, h, w )\n * elif type(pyobj) == list and len(pyobj[0]) > 4:\n */\n __pyx_t_1 = PyObject_RichCompare(((PyObject *)Py_TYPE(__pyx_v_pyobj)), ((PyObject *)(&PyList_Type)), Py_EQ); __Pyx_XGOTREF(__pyx_t_1); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 283, __pyx_L1_error)\n __pyx_t_8 = __Pyx_PyObject_IsTrue(__pyx_t_1); if (unlikely(__pyx_t_8 < 0)) __PYX_ERR(0, 283, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n if (__pyx_t_8) {\n } else {\n __pyx_t_2 = __pyx_t_8;\n goto __pyx_L4_bool_binop_done;\n }\n __pyx_t_1 = __Pyx_GetItemInt(__pyx_v_pyobj, 0, long, 1, __Pyx_PyInt_From_long, 0, 0, 1); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 283, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_9 = PyObject_Length(__pyx_t_1); if (unlikely(__pyx_t_9 == ((Py_ssize_t)-1))) __PYX_ERR(0, 283, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __pyx_t_8 = ((__pyx_t_9 == 4) != 0);\n __pyx_t_2 = __pyx_t_8;\n __pyx_L4_bool_binop_done:;\n if (__pyx_t_2) {\n\n /* \"pycocotools/_mask.pyx\":284\n * objs = frBbox(pyobj, h, w )\n * elif type(pyobj) == list and len(pyobj[0]) == 4:\n * objs = frBbox(pyobj, h, w ) # <<<<<<<<<<<<<<\n * elif type(pyobj) == list and len(pyobj[0]) > 4:\n * objs = frPoly(pyobj, h, w )\n */\n __Pyx_GetModuleGlobalName(__pyx_t_3, __pyx_n_s_frBbox); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 284, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __pyx_t_7 = __Pyx_PyInt_From_siz(__pyx_v_h); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 284, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_7);\n __pyx_t_4 = NULL;\n __pyx_t_6 = 0;\n if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_3))) {\n __pyx_t_4 = PyMethod_GET_SELF(__pyx_t_3);\n if (likely(__pyx_t_4)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_3);\n __Pyx_INCREF(__pyx_t_4);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_3, function);\n __pyx_t_6 = 1;\n }\n }\n #if CYTHON_FAST_PYCALL\n if (PyFunction_Check(__pyx_t_3)) {\n PyObject *__pyx_temp[4] = {__pyx_t_4, __pyx_v_pyobj, __pyx_t_7, __pyx_v_w};\n __pyx_t_1 = __Pyx_PyFunction_FastCall(__pyx_t_3, __pyx_temp+1-__pyx_t_6, 3+__pyx_t_6); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 284, __pyx_L1_error)\n __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;\n } else\n #endif\n #if CYTHON_FAST_PYCCALL\n if (__Pyx_PyFastCFunction_Check(__pyx_t_3)) {\n PyObject *__pyx_temp[4] = {__pyx_t_4, __pyx_v_pyobj, __pyx_t_7, __pyx_v_w};\n __pyx_t_1 = __Pyx_PyCFunction_FastCall(__pyx_t_3, __pyx_temp+1-__pyx_t_6, 3+__pyx_t_6); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 284, __pyx_L1_error)\n __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;\n } else\n #endif\n {\n __pyx_t_5 = PyTuple_New(3+__pyx_t_6); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 284, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n if (__pyx_t_4) {\n __Pyx_GIVEREF(__pyx_t_4); PyTuple_SET_ITEM(__pyx_t_5, 0, __pyx_t_4); __pyx_t_4 = NULL;\n }\n __Pyx_INCREF(__pyx_v_pyobj);\n __Pyx_GIVEREF(__pyx_v_pyobj);\n PyTuple_SET_ITEM(__pyx_t_5, 0+__pyx_t_6, __pyx_v_pyobj);\n __Pyx_GIVEREF(__pyx_t_7);\n PyTuple_SET_ITEM(__pyx_t_5, 1+__pyx_t_6, __pyx_t_7);\n __Pyx_INCREF(__pyx_v_w);\n __Pyx_GIVEREF(__pyx_v_w);\n PyTuple_SET_ITEM(__pyx_t_5, 2+__pyx_t_6, __pyx_v_w);\n __pyx_t_7 = 0;\n __pyx_t_1 = __Pyx_PyObject_Call(__pyx_t_3, __pyx_t_5, NULL); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 284, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n }\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __pyx_v_objs = __pyx_t_1;\n __pyx_t_1 = 0;\n\n /* \"pycocotools/_mask.pyx\":283\n * if type(pyobj) == np.ndarray:\n * objs = frBbox(pyobj, h, w )\n * elif type(pyobj) == list and len(pyobj[0]) == 4: # <<<<<<<<<<<<<<\n * objs = frBbox(pyobj, h, w )\n * elif type(pyobj) == list and len(pyobj[0]) > 4:\n */\n goto __pyx_L3;\n }\n\n /* \"pycocotools/_mask.pyx\":285\n * elif type(pyobj) == list and len(pyobj[0]) == 4:\n * objs = frBbox(pyobj, h, w )\n * elif type(pyobj) == list and len(pyobj[0]) > 4: # <<<<<<<<<<<<<<\n * objs = frPoly(pyobj, h, w )\n * elif type(pyobj) == list and type(pyobj[0]) == dict:\n */\n __pyx_t_1 = PyObject_RichCompare(((PyObject *)Py_TYPE(__pyx_v_pyobj)), ((PyObject *)(&PyList_Type)), Py_EQ); __Pyx_XGOTREF(__pyx_t_1); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 285, __pyx_L1_error)\n __pyx_t_8 = __Pyx_PyObject_IsTrue(__pyx_t_1); if (unlikely(__pyx_t_8 < 0)) __PYX_ERR(0, 285, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n if (__pyx_t_8) {\n } else {\n __pyx_t_2 = __pyx_t_8;\n goto __pyx_L6_bool_binop_done;\n }\n __pyx_t_1 = __Pyx_GetItemInt(__pyx_v_pyobj, 0, long, 1, __Pyx_PyInt_From_long, 0, 0, 1); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 285, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_9 = PyObject_Length(__pyx_t_1); if (unlikely(__pyx_t_9 == ((Py_ssize_t)-1))) __PYX_ERR(0, 285, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __pyx_t_8 = ((__pyx_t_9 > 4) != 0);\n __pyx_t_2 = __pyx_t_8;\n __pyx_L6_bool_binop_done:;\n if (__pyx_t_2) {\n\n /* \"pycocotools/_mask.pyx\":286\n * objs = frBbox(pyobj, h, w )\n * elif type(pyobj) == list and len(pyobj[0]) > 4:\n * objs = frPoly(pyobj, h, w ) # <<<<<<<<<<<<<<\n * elif type(pyobj) == list and type(pyobj[0]) == dict:\n * objs = frUncompressedRLE(pyobj, h, w)\n */\n __Pyx_GetModuleGlobalName(__pyx_t_3, __pyx_n_s_frPoly); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 286, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __pyx_t_5 = __Pyx_PyInt_From_siz(__pyx_v_h); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 286, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_5);\n __pyx_t_7 = NULL;\n __pyx_t_6 = 0;\n if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_3))) {\n __pyx_t_7 = PyMethod_GET_SELF(__pyx_t_3);\n if (likely(__pyx_t_7)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_3);\n __Pyx_INCREF(__pyx_t_7);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_3, function);\n __pyx_t_6 = 1;\n }\n }\n #if CYTHON_FAST_PYCALL\n if (PyFunction_Check(__pyx_t_3)) {\n PyObject *__pyx_temp[4] = {__pyx_t_7, __pyx_v_pyobj, __pyx_t_5, __pyx_v_w};\n __pyx_t_1 = __Pyx_PyFunction_FastCall(__pyx_t_3, __pyx_temp+1-__pyx_t_6, 3+__pyx_t_6); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 286, __pyx_L1_error)\n __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n } else\n #endif\n #if CYTHON_FAST_PYCCALL\n if (__Pyx_PyFastCFunction_Check(__pyx_t_3)) {\n PyObject *__pyx_temp[4] = {__pyx_t_7, __pyx_v_pyobj, __pyx_t_5, __pyx_v_w};\n __pyx_t_1 = __Pyx_PyCFunction_FastCall(__pyx_t_3, __pyx_temp+1-__pyx_t_6, 3+__pyx_t_6); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 286, __pyx_L1_error)\n __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;\n } else\n #endif\n {\n __pyx_t_4 = PyTuple_New(3+__pyx_t_6); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 286, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n if (__pyx_t_7) {\n __Pyx_GIVEREF(__pyx_t_7); PyTuple_SET_ITEM(__pyx_t_4, 0, __pyx_t_7); __pyx_t_7 = NULL;\n }\n __Pyx_INCREF(__pyx_v_pyobj);\n __Pyx_GIVEREF(__pyx_v_pyobj);\n PyTuple_SET_ITEM(__pyx_t_4, 0+__pyx_t_6, __pyx_v_pyobj);\n __Pyx_GIVEREF(__pyx_t_5);\n PyTuple_SET_ITEM(__pyx_t_4, 1+__pyx_t_6, __pyx_t_5);\n __Pyx_INCREF(__pyx_v_w);\n __Pyx_GIVEREF(__pyx_v_w);\n PyTuple_SET_ITEM(__pyx_t_4, 2+__pyx_t_6, __pyx_v_w);\n __pyx_t_5 = 0;\n __pyx_t_1 = __Pyx_PyObject_Call(__pyx_t_3, __pyx_t_4, NULL); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 286, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;\n }\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __pyx_v_objs = __pyx_t_1;\n __pyx_t_1 = 0;\n\n /* \"pycocotools/_mask.pyx\":285\n * elif type(pyobj) == list and len(pyobj[0]) == 4:\n * objs = frBbox(pyobj, h, w )\n * elif type(pyobj) == list and len(pyobj[0]) > 4: # <<<<<<<<<<<<<<\n * objs = frPoly(pyobj, h, w )\n * elif type(pyobj) == list and type(pyobj[0]) == dict:\n */\n goto __pyx_L3;\n }\n\n /* \"pycocotools/_mask.pyx\":287\n * elif type(pyobj) == list and len(pyobj[0]) > 4:\n * objs = frPoly(pyobj, h, w )\n * elif type(pyobj) == list and type(pyobj[0]) == dict: # <<<<<<<<<<<<<<\n * objs = frUncompressedRLE(pyobj, h, w)\n * else:\n */\n __pyx_t_1 = PyObject_RichCompare(((PyObject *)Py_TYPE(__pyx_v_pyobj)), ((PyObject *)(&PyList_Type)), Py_EQ); __Pyx_XGOTREF(__pyx_t_1); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 287, __pyx_L1_error)\n __pyx_t_8 = __Pyx_PyObject_IsTrue(__pyx_t_1); if (unlikely(__pyx_t_8 < 0)) __PYX_ERR(0, 287, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n if (__pyx_t_8) {\n } else {\n __pyx_t_2 = __pyx_t_8;\n goto __pyx_L8_bool_binop_done;\n }\n __pyx_t_1 = __Pyx_GetItemInt(__pyx_v_pyobj, 0, long, 1, __Pyx_PyInt_From_long, 0, 0, 1); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 287, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_3 = PyObject_RichCompare(((PyObject *)Py_TYPE(__pyx_t_1)), ((PyObject *)(&PyDict_Type)), Py_EQ); __Pyx_XGOTREF(__pyx_t_3); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 287, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __pyx_t_8 = __Pyx_PyObject_IsTrue(__pyx_t_3); if (unlikely(__pyx_t_8 < 0)) __PYX_ERR(0, 287, __pyx_L1_error)\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __pyx_t_2 = __pyx_t_8;\n __pyx_L8_bool_binop_done:;\n if (likely(__pyx_t_2)) {\n\n /* \"pycocotools/_mask.pyx\":288\n * objs = frPoly(pyobj, h, w )\n * elif type(pyobj) == list and type(pyobj[0]) == dict:\n * objs = frUncompressedRLE(pyobj, h, w) # <<<<<<<<<<<<<<\n * else:\n * raise Exception('input type is not supported.')\n */\n __Pyx_GetModuleGlobalName(__pyx_t_1, __pyx_n_s_frUncompressedRLE); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 288, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_1);\n __pyx_t_4 = __Pyx_PyInt_From_siz(__pyx_v_h); if (unlikely(!__pyx_t_4)) __PYX_ERR(0, 288, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_4);\n __pyx_t_5 = NULL;\n __pyx_t_6 = 0;\n if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_1))) {\n __pyx_t_5 = PyMethod_GET_SELF(__pyx_t_1);\n if (likely(__pyx_t_5)) {\n PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_1);\n __Pyx_INCREF(__pyx_t_5);\n __Pyx_INCREF(function);\n __Pyx_DECREF_SET(__pyx_t_1, function);\n __pyx_t_6 = 1;\n }\n }\n #if CYTHON_FAST_PYCALL\n if (PyFunction_Check(__pyx_t_1)) {\n PyObject *__pyx_temp[4] = {__pyx_t_5, __pyx_v_pyobj, __pyx_t_4, __pyx_v_w};\n __pyx_t_3 = __Pyx_PyFunction_FastCall(__pyx_t_1, __pyx_temp+1-__pyx_t_6, 3+__pyx_t_6); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 288, __pyx_L1_error)\n __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;\n } else\n #endif\n #if CYTHON_FAST_PYCCALL\n if (__Pyx_PyFastCFunction_Check(__pyx_t_1)) {\n PyObject *__pyx_temp[4] = {__pyx_t_5, __pyx_v_pyobj, __pyx_t_4, __pyx_v_w};\n __pyx_t_3 = __Pyx_PyCFunction_FastCall(__pyx_t_1, __pyx_temp+1-__pyx_t_6, 3+__pyx_t_6); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 288, __pyx_L1_error)\n __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_DECREF(__pyx_t_4); __pyx_t_4 = 0;\n } else\n #endif\n {\n __pyx_t_7 = PyTuple_New(3+__pyx_t_6); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 288, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_7);\n if (__pyx_t_5) {\n __Pyx_GIVEREF(__pyx_t_5); PyTuple_SET_ITEM(__pyx_t_7, 0, __pyx_t_5); __pyx_t_5 = NULL;\n }\n __Pyx_INCREF(__pyx_v_pyobj);\n __Pyx_GIVEREF(__pyx_v_pyobj);\n PyTuple_SET_ITEM(__pyx_t_7, 0+__pyx_t_6, __pyx_v_pyobj);\n __Pyx_GIVEREF(__pyx_t_4);\n PyTuple_SET_ITEM(__pyx_t_7, 1+__pyx_t_6, __pyx_t_4);\n __Pyx_INCREF(__pyx_v_w);\n __Pyx_GIVEREF(__pyx_v_w);\n PyTuple_SET_ITEM(__pyx_t_7, 2+__pyx_t_6, __pyx_v_w);\n __pyx_t_4 = 0;\n __pyx_t_3 = __Pyx_PyObject_Call(__pyx_t_1, __pyx_t_7, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 288, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;\n }\n __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;\n __pyx_v_objs = __pyx_t_3;\n __pyx_t_3 = 0;\n\n /* \"pycocotools/_mask.pyx\":287\n * elif type(pyobj) == list and len(pyobj[0]) > 4:\n * objs = frPoly(pyobj, h, w )\n * elif type(pyobj) == list and type(pyobj[0]) == dict: # <<<<<<<<<<<<<<\n * objs = frUncompressedRLE(pyobj, h, w)\n * else:\n */\n goto __pyx_L3;\n }\n\n /* \"pycocotools/_mask.pyx\":290\n * objs = frUncompressedRLE(pyobj, h, w)\n * else:\n * raise Exception('input type is not supported.') # <<<<<<<<<<<<<<\n * return objs\n */\n /*else*/ {\n __pyx_t_3 = __Pyx_PyObject_Call(((PyObject *)(&((PyTypeObject*)PyExc_Exception)[0])), __pyx_tuple__19, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 290, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_Raise(__pyx_t_3, 0, 0, 0);\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __PYX_ERR(0, 290, __pyx_L1_error)\n }\n __pyx_L3:;\n\n /* \"pycocotools/_mask.pyx\":291\n * else:\n * raise Exception('input type is not supported.')\n * return objs # <<<<<<<<<<<<<<\n */\n __Pyx_XDECREF(__pyx_r);\n __Pyx_INCREF(__pyx_v_objs);\n __pyx_r = __pyx_v_objs;\n goto __pyx_L0;\n\n /* \"pycocotools/_mask.pyx\":280\n * return objs\n * \n * def frPyObjects(pyobj, siz h, w): # <<<<<<<<<<<<<<\n * if type(pyobj) == np.ndarray:\n * objs = frBbox(pyobj, h, w )\n */\n\n /* function exit code */\n __pyx_L1_error:;\n __Pyx_XDECREF(__pyx_t_1);\n __Pyx_XDECREF(__pyx_t_3);\n __Pyx_XDECREF(__pyx_t_4);\n __Pyx_XDECREF(__pyx_t_5);\n __Pyx_XDECREF(__pyx_t_7);\n __Pyx_AddTraceback(\"pycocotools._mask.frPyObjects\", __pyx_clineno, __pyx_lineno, __pyx_filename);\n __pyx_r = NULL;\n __pyx_L0:;\n __Pyx_XDECREF(__pyx_v_objs);\n __Pyx_XGIVEREF(__pyx_r);\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\n/* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":258\n * # experimental exception made for __getbuffer__ and __releasebuffer__\n * # -- the details of this may change.\n * def __getbuffer__(ndarray self, Py_buffer* info, int flags): # <<<<<<<<<<<<<<\n * # This implementation of getbuffer is geared towards Cython\n * # requirements, and does not yet fulfill the PEP.\n */\n\n/* Python wrapper */\nstatic CYTHON_UNUSED int __pyx_pw_5numpy_7ndarray_1__getbuffer__(PyObject *__pyx_v_self, Py_buffer *__pyx_v_info, int __pyx_v_flags); /*proto*/\nstatic CYTHON_UNUSED int __pyx_pw_5numpy_7ndarray_1__getbuffer__(PyObject *__pyx_v_self, Py_buffer *__pyx_v_info, int __pyx_v_flags) {\n int __pyx_r;\n __Pyx_RefNannyDeclarations\n __Pyx_RefNannySetupContext(\"__getbuffer__ (wrapper)\", 0);\n __pyx_r = __pyx_pf_5numpy_7ndarray___getbuffer__(((PyArrayObject *)__pyx_v_self), ((Py_buffer *)__pyx_v_info), ((int)__pyx_v_flags));\n\n /* function exit code */\n __Pyx_RefNannyFinishContext();\n return __pyx_r;\n}\n\nstatic int __pyx_pf_5numpy_7ndarray___getbuffer__(PyArrayObject *__pyx_v_self, Py_buffer *__pyx_v_info, int __pyx_v_flags) {\n int __pyx_v_i;\n int __pyx_v_ndim;\n int __pyx_v_endian_detector;\n int __pyx_v_little_endian;\n int __pyx_v_t;\n char *__pyx_v_f;\n PyArray_Descr *__pyx_v_descr = 0;\n int __pyx_v_offset;\n int __pyx_r;\n __Pyx_RefNannyDeclarations\n int __pyx_t_1;\n int __pyx_t_2;\n PyObject *__pyx_t_3 = NULL;\n int __pyx_t_4;\n int __pyx_t_5;\n int __pyx_t_6;\n PyArray_Descr *__pyx_t_7;\n PyObject *__pyx_t_8 = NULL;\n char *__pyx_t_9;\n if (__pyx_v_info == NULL) {\n PyErr_SetString(PyExc_BufferError, \"PyObject_GetBuffer: view==NULL argument is obsolete\");\n return -1;\n }\n __Pyx_RefNannySetupContext(\"__getbuffer__\", 0);\n __pyx_v_info->obj = Py_None; __Pyx_INCREF(Py_None);\n __Pyx_GIVEREF(__pyx_v_info->obj);\n\n /* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":265\n * \n * cdef int i, ndim\n * cdef int endian_detector = 1 # <<<<<<<<<<<<<<\n * cdef bint little_endian = ((&endian_detector)[0] != 0)\n * \n */\n __pyx_v_endian_detector = 1;\n\n /* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":266\n * cdef int i, ndim\n * cdef int endian_detector = 1\n * cdef bint little_endian = ((&endian_detector)[0] != 0) # <<<<<<<<<<<<<<\n * \n * ndim = PyArray_NDIM(self)\n */\n __pyx_v_little_endian = ((((char *)(&__pyx_v_endian_detector))[0]) != 0);\n\n /* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":268\n * cdef bint little_endian = ((&endian_detector)[0] != 0)\n * \n * ndim = PyArray_NDIM(self) # <<<<<<<<<<<<<<\n * \n * if ((flags & pybuf.PyBUF_C_CONTIGUOUS == pybuf.PyBUF_C_CONTIGUOUS)\n */\n __pyx_v_ndim = PyArray_NDIM(__pyx_v_self);\n\n /* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":270\n * ndim = PyArray_NDIM(self)\n * \n * if ((flags & pybuf.PyBUF_C_CONTIGUOUS == pybuf.PyBUF_C_CONTIGUOUS) # <<<<<<<<<<<<<<\n * and not PyArray_CHKFLAGS(self, NPY_ARRAY_C_CONTIGUOUS)):\n * raise ValueError(u\"ndarray is not C contiguous\")\n */\n __pyx_t_2 = (((__pyx_v_flags & PyBUF_C_CONTIGUOUS) == PyBUF_C_CONTIGUOUS) != 0);\n if (__pyx_t_2) {\n } else {\n __pyx_t_1 = __pyx_t_2;\n goto __pyx_L4_bool_binop_done;\n }\n\n /* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":271\n * \n * if ((flags & pybuf.PyBUF_C_CONTIGUOUS == pybuf.PyBUF_C_CONTIGUOUS)\n * and not PyArray_CHKFLAGS(self, NPY_ARRAY_C_CONTIGUOUS)): # <<<<<<<<<<<<<<\n * raise ValueError(u\"ndarray is not C contiguous\")\n * \n */\n __pyx_t_2 = ((!(PyArray_CHKFLAGS(__pyx_v_self, NPY_ARRAY_C_CONTIGUOUS) != 0)) != 0);\n __pyx_t_1 = __pyx_t_2;\n __pyx_L4_bool_binop_done:;\n\n /* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":270\n * ndim = PyArray_NDIM(self)\n * \n * if ((flags & pybuf.PyBUF_C_CONTIGUOUS == pybuf.PyBUF_C_CONTIGUOUS) # <<<<<<<<<<<<<<\n * and not PyArray_CHKFLAGS(self, NPY_ARRAY_C_CONTIGUOUS)):\n * raise ValueError(u\"ndarray is not C contiguous\")\n */\n if (unlikely(__pyx_t_1)) {\n\n /* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":272\n * if ((flags & pybuf.PyBUF_C_CONTIGUOUS == pybuf.PyBUF_C_CONTIGUOUS)\n * and not PyArray_CHKFLAGS(self, NPY_ARRAY_C_CONTIGUOUS)):\n * raise ValueError(u\"ndarray is not C contiguous\") # <<<<<<<<<<<<<<\n * \n * if ((flags & pybuf.PyBUF_F_CONTIGUOUS == pybuf.PyBUF_F_CONTIGUOUS)\n */\n __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_ValueError, __pyx_tuple__20, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(2, 272, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_Raise(__pyx_t_3, 0, 0, 0);\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __PYX_ERR(2, 272, __pyx_L1_error)\n\n /* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":270\n * ndim = PyArray_NDIM(self)\n * \n * if ((flags & pybuf.PyBUF_C_CONTIGUOUS == pybuf.PyBUF_C_CONTIGUOUS) # <<<<<<<<<<<<<<\n * and not PyArray_CHKFLAGS(self, NPY_ARRAY_C_CONTIGUOUS)):\n * raise ValueError(u\"ndarray is not C contiguous\")\n */\n }\n\n /* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":274\n * raise ValueError(u\"ndarray is not C contiguous\")\n * \n * if ((flags & pybuf.PyBUF_F_CONTIGUOUS == pybuf.PyBUF_F_CONTIGUOUS) # <<<<<<<<<<<<<<\n * and not PyArray_CHKFLAGS(self, NPY_ARRAY_F_CONTIGUOUS)):\n * raise ValueError(u\"ndarray is not Fortran contiguous\")\n */\n __pyx_t_2 = (((__pyx_v_flags & PyBUF_F_CONTIGUOUS) == PyBUF_F_CONTIGUOUS) != 0);\n if (__pyx_t_2) {\n } else {\n __pyx_t_1 = __pyx_t_2;\n goto __pyx_L7_bool_binop_done;\n }\n\n /* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":275\n * \n * if ((flags & pybuf.PyBUF_F_CONTIGUOUS == pybuf.PyBUF_F_CONTIGUOUS)\n * and not PyArray_CHKFLAGS(self, NPY_ARRAY_F_CONTIGUOUS)): # <<<<<<<<<<<<<<\n * raise ValueError(u\"ndarray is not Fortran contiguous\")\n * \n */\n __pyx_t_2 = ((!(PyArray_CHKFLAGS(__pyx_v_self, NPY_ARRAY_F_CONTIGUOUS) != 0)) != 0);\n __pyx_t_1 = __pyx_t_2;\n __pyx_L7_bool_binop_done:;\n\n /* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":274\n * raise ValueError(u\"ndarray is not C contiguous\")\n * \n * if ((flags & pybuf.PyBUF_F_CONTIGUOUS == pybuf.PyBUF_F_CONTIGUOUS) # <<<<<<<<<<<<<<\n * and not PyArray_CHKFLAGS(self, NPY_ARRAY_F_CONTIGUOUS)):\n * raise ValueError(u\"ndarray is not Fortran contiguous\")\n */\n if (unlikely(__pyx_t_1)) {\n\n /* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":276\n * if ((flags & pybuf.PyBUF_F_CONTIGUOUS == pybuf.PyBUF_F_CONTIGUOUS)\n * and not PyArray_CHKFLAGS(self, NPY_ARRAY_F_CONTIGUOUS)):\n * raise ValueError(u\"ndarray is not Fortran contiguous\") # <<<<<<<<<<<<<<\n * \n * info.buf = PyArray_DATA(self)\n */\n __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_ValueError, __pyx_tuple__21, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(2, 276, __pyx_L1_error)\n __Pyx_GOTREF(__pyx_t_3);\n __Pyx_Raise(__pyx_t_3, 0, 0, 0);\n __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;\n __PYX_ERR(2, 276, __pyx_L1_error)\n\n /* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":274\n * raise ValueError(u\"ndarray is not C contiguous\")\n * \n * if ((flags & pybuf.PyBUF_F_CONTIGUOUS == pybuf.PyBUF_F_CONTIGUOUS) # <<<<<<<<<<<<<<\n * and not PyArray_CHKFLAGS(self, NPY_ARRAY_F_CONTIGUOUS)):\n * raise ValueError(u\"ndarray is not Fortran contiguous\")\n */\n }\n\n /* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":278\n * raise ValueError(u\"ndarray is not Fortran contiguous\")\n * \n * info.buf = PyArray_DATA(self) # <<<<<<<<<<<<<<\n * info.ndim = ndim\n * if sizeof(npy_intp) != sizeof(Py_ssize_t):\n */\n __pyx_v_info->buf = PyArray_DATA(__pyx_v_self);\n\n /* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":279\n * \n * info.buf = PyArray_DATA(self)\n * info.ndim = ndim # <<<<<<<<<<<<<<\n * if sizeof(npy_intp) != sizeof(Py_ssize_t):\n * # Allocate new buffer for strides and shape info.\n */\n __pyx_v_info->ndim = __pyx_v_ndim;\n\n /* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":280\n * info.buf = PyArray_DATA(self)\n * info.ndim = ndim\n * if sizeof(npy_intp) != sizeof(Py_ssize_t): # <<<<<<<<<<<<<<\n * # Allocate new buffer for strides and shape info.\n * # This is allocated as one block, strides first.\n */\n __pyx_t_1 = (((sizeof(npy_intp)) != (sizeof(Py_ssize_t))) != 0);\n if (__pyx_t_1) {\n\n /* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":283\n * # Allocate new buffer for strides and shape info.\n * # This is allocated as one block, strides first.\n * info.strides = PyObject_Malloc(sizeof(Py_ssize_t) * 2 * ndim) # <<<<<<<<<<<<<<\n * info.shape = info.strides + ndim\n * for i in range(ndim):\n */\n __pyx_v_info->strides = ((Py_ssize_t *)PyObject_Malloc((((sizeof(Py_ssize_t)) * 2) * ((size_t)__pyx_v_ndim))));\n\n /* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":284\n * # This is allocated as one block, strides first.\n * info.strides = PyObject_Malloc(sizeof(Py_ssize_t) * 2 * ndim)\n * info.shape = info.strides + ndim # <<<<<<<<<<<<<<\n * for i in range(ndim):\n * info.strides[i] = PyArray_STRIDES(self)[i]\n */\n __pyx_v_info->shape = (__pyx_v_info->strides + __pyx_v_ndim);\n\n /* \"../../../anaconda3/envs/pytorch0.3/lib/python3.6/site-packages/Cython/Includes/numpy/__init__.pxd\":285\n * info.strides = PyObject_Malloc(sizeof(Py_ssize_t) * 2 * | |